PROBLEMS

4th
Edition

AND

SOLUTIONS IN
QUANTUM
COMPUTING
AND
QUANTUM
INFORMATION

10943hc_9789813238404_tp.indd 1 30/1/18 2:00 PM

August 1, 2014 10:17 9in x 6in Scanning Transmission Electron Microscopy. . . b1683-fm

This page intentionally left blank

 Willi-Hans Steeb
University of Johannesburg, South Africa

Yorick Hardy
University of the Witwatersrand, South Africa

PROBLEMS 4th
Edition

AND

SOLUTIONS IN
QUANTUM
COMPUTING
AND
QUANTUM
INFORMATION

World Scientific
NEW JERSEY • LONDON • SINGAPORE • BEIJING • SHANGHAI • HONG KONG • TA I P E I • CHENNAI

10943hc_9789813238404_tp.indd 2 30/1/18 2:00 PM

 Published by
World Scientific Publishing Co. Pte. Ltd.
5 Toh Tuck Link, Singapore 596224
USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601
UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE

Library of Congress Cataloging-in-Publication Data

Names: Steeb, W.-H., author. | Hardy, Yorick, 1976– author.
Title: Problems and solutions in quantum computing and quantum information /
Willi-Hans Steeb and Yorick Hardy, University of Johannesburg, South Africa.
Description: 4th edition. | Singapore ; Hackensack, NJ : World Scientific, [2018] |
Includes bibliographical references and index.
Identifiers: LCCN 2018001917| ISBN 9789813238404 (hardcover ; alk. paper) |
ISBN 9813238402 (hardcover ; alk. paper) | ISBN 9789813239289 (pbk ; alk. paper) |
ISBN 981323928X (pbk ; alk. paper)
Subjects: LCSH: Quantum computers. | Quantum theory.
Classification: LCC QA76.889 .S74 2018 | DDC 006.3/843--dc23
LC record available at https://lccn.loc.gov/2018001917

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

Copyright © 2018 by World Scientific Publishing Co. Pte. Ltd.

All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means,
electronic or mechanical, including photocopying, recording or any information storage and retrieval
system now known or to be invented, without written permission from the publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance
Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy
is not required from the publisher.

For any available supplementary material, please visit

http://www.worldscientific.com/worldscibooks/10.1142/10943#t=suppl

Printed in Singapore

Lakshmi - 10943 - Problems and Solutions in Quantum Computing.indd 1 12-03-18 2:59:58 PM

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page v

Preface
The purpose of this book is to supply a collection of problems in quan-
tum computing and quantum information together with their detailed so-
lutions which will prove to be valuable to graduate students as well as to
research workers in these fields. All the important concepts and topics such
as quantum gates and quantum circuits, quantum channels, entanglement,
teleportation, Bell states, Bell inequality, Schmidt decomposition, quantum
Fourier transform, magic gate, von Neumann entropy, quantum channels,
quantum cryptography, quantum error correction, coherent states, coherent
Bell states, squeezed states, POVM measurement, beam splitter, homodyne
detection and Kerr Hamilton operator are included. The topics range in
difficulty from elementary to advanced. Almost all problems are solved
in detail and most of the problems are self-contained. All relevant defi-
nitions are given. Students can learn important principles and strategies
required for problem solving. Teachers will also find this text useful as a
supplement, since important concepts and techniques are developed in the
problems. The book can also be used as a text or a supplement for linear
and multilinear algebra or matrix theory. Each chapter also includes sup-
plementary problems.

Most chapters also include programming problems in Maxima and Symbol-

icC++.

The material was tested in our lectures given around the world.

Any useful suggestions and comments are welcome.

The International School for Scientific Computing (ISSC) provides certifi-

cate courses for this subject. Please contact the first author if you want
to do this course. More quantum computing exercises can be found on the
web page given below.

e-mail addresses of the authors:

[email protected]
[email protected]
[email protected]
[email protected]
Home page of the authors: http://issc.uj.ac.za

v
August 1, 2014 10:17 9in x 6in Scanning Transmission Electron Microscopy. . . b1683-fm

This page intentionally left blank

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page vii

Contents
Preface v

Notation xi

I Finite-Dimensional Hilbert Spaces 1

1 Qubits 3
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 27

2 Kronecker Product and Tensor Product 31

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 33
2.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 59

3 Matrix Properties 65
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 67
3.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 108

4 Density Operators 115

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 117
4.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 139

5 Trace and Partial Trace 141

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
5.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 143
5.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 153

6 Boolean Functions and Quantum Gates 155

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
6.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 157
6.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 173

vii
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page viii

viii Contents

7 Unitary Transforms and Quantum Gates 175

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
7.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 177
7.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 208

8 Entropy 209
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
8.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 211
8.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 225

9 Measurement 227
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
9.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 229
9.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 249

10 Entanglement 251
10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
10.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 253
10.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 293

11 Bell Inequality 295

11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
11.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 297
11.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 306

12 Teleportation 307
12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
12.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 308
12.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 318

13 Cloning 319
13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
13.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 320
13.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 324

14 Quantum Algorithms 325

14.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
14.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 327
14.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 349

15 Quantum Error Correction 351

15.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
15.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 353
15.3 Supplementary Problem . . . . . . . . . . . . . . . . . . . . 358
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page ix

Contents ix

16 Quantum Cryptography 359

16.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
16.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 360
16.3 Supplementary Problem . . . . . . . . . . . . . . . . . . . . 368

17 Quantum Channels 369

17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
17.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 371
17.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 384

II Infinite-Dimensional Hilbert Spaces 385

18 Bose Operators and Number States 387
18.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
18.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 389
18.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 417

19 Coherent States 421

19.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
19.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 423
19.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 437

20 Squeezed States 439

20.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
20.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 441
20.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 457

21 Trace and Partial Trace 461

21.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 461
21.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 462
21.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 468

22 Entanglement 469
22.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
22.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 471
22.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 487

23 Continuous Variable Teleportation 489

23.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
23.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 489
23.3 Supplementary Problem . . . . . . . . . . . . . . . . . . . . 493
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page x

x Contents

24 Swapping and Cloning 495

24.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
24.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 495
24.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 500

25 Homodyne Detection 501

25.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
25.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 502
25.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 504

26 Hamilton Operators 505

26.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
26.2 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . . 506
26.3 Supplementary Problems . . . . . . . . . . . . . . . . . . . . 519

Bibliography 521

Index 535
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page xi

Notation
∅ empty set
N natural numbers
N0 natural numbers including 0
Z integers
Q rational numbers
R real numbers
R+ nonnegative real numbers
C complex numbers
Rn n-dimensional Euclidean space
Cn n-dimensional complex linear space
H Hilbert
√ space
i −1
ωn := exp(2πi/n) n-th root of unity n ∈ N
<(z) real part of the complex number z
=(z) imaginary part of the complex number z
A⊂B subset A of set B
A∩B the intersection of the sets A and B
A∪B the union of the sets A and B
f ◦g composition of two mappings (f ◦ g)(x) = f (g(x))
x column vector in Cn
xT transpose of x (row vector)
× vector product in R3
k.k norm
x · y ≡ x∗ y scalar product (inner product) in Cn
h.|.i scalar product in Hilbert space
x×y vector product
A⊕B direct sum of matrices A and B
A⊗B Kronecker product of matrices A and B
f ⊗g tensor product of elements f and g
of Hilbert spaces
det(A) determinant of a square matrix A
tr(A) trace of a square matrix A
rank(A) rank of matrix A
AT transpose of matrix A
A conjugate of matrix A
A∗ conjugate transpose of matrix A
A† conjugate transpose of matrix A
(notation used in physics)

xi
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page xii

xii Notation

In n × n unit matrix
I unit operator
U unitary operator, unitary matrix
Π projection operator, projection matrix
P permutation matrix
σ1 , σ2 , σ3 Pauli spin matrices
S1 , S2 , S3 spin matrices for spin s = 1/2, 1, 3/2, 2, . . .
ρ density operator, density matrix
[A, B] := AB − BA commutator for square matrices A and B
[A, B]+ := AB + BA anticommutator for square matrices A and B
δjk Kronecker delta with δjk = 1 for j = k
and δjk = 0 for j 6= k
Ejk elementary matrices with 1 at jk and 0 otherwise
λ eigenvalue
 real parameter
H Hamilton function
Ĥ Hamilton operator
{|0i, |1i, . . . , |n − 1i } arbitrary orthonormal basis for Cn
|Φ+ i, |Φ− i, |Ψ+ i, |Ψ− i Bell states in C4
~ h/2π with h the Planck constant
t time
ω frequency
k wave vector
r space coordinates
p momentum
b, b† Bose annihilation and creation operators
|ni number states, Fock states n = 0, 1, 2, . . .
|βi coherent state, β ∈ C
|ζi squeezed state, ζ ∈ C
D(β) displacement operator, β ∈ C
S(ζ) one-mode squeezing operator, ζ ∈ C
s squeezing parameter
c, c† Fermi annihilation and creation operators
E electric field
B magnetic induction
P electric polarization

The Pauli spin matrices are used extensively in the book. They are given
by
     
0 1 0 −i 1 0
σ1 := , σ2 := , σ3 := .
1 0 i 0 0 −1

The spin- 21 matrix S1 , S2 , S3 are given by S1 = 12 σ1 , S2 = 12 σ2 , S3 = 12 σ3 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page xiii

Notation xiii

The Dirac gamma matrices are given by

γ1 = −σ2 ⊗ σ1 , γ2 = σ2 ⊗ σ2 , γ3 = σ2 ⊗ σ3 , γ4 = σ3 ⊗ I2 .

The concept of a Hilbert space will be used throughout the book. A Hilbert
space is a set, H of elements, or vectors, (f, g, h, . . .) which satisfies the
following conditions (1) - - (5).

(1) If f and g belong to H, then there is a unique element of H, denoted

by f + g, the operation of addition (+) being invertible, commutative and
associative.

(2) If c is a complex number, then for any f in H, there is an element cf of

H; and the multiplication of vectors by complex numbers thereby defined
satisfies the distributive conditions

c(f + g) = cf + cg, (c1 + c2 )f = c1 f + c2 f.

(3) Hilbert spaces H possess a zero element, 0, characterized by the prop-

erty that 0 + f = f for all vectors f in H.

(4) For each pair of vectors f , g in H, there is a complex number hf |gi,

termed the inner product or scalar product, of f with g, such that

hf |gi = hg|f i

hf |g + hi = hf |gi + hf |hi
hf |cgi = chf |gi
and
hf |f i ≥ 0.
Equality in the last formula occurs only if f = 0. The scalar product defines
the norm kf k = hf |f i1/2 .

(5) If { fn } is a sequence in H satisfying the Cauchy condition that

kfm − fn k → 0

as m and n tend independently to infinity, then there is a unique element

f of H such that kfn − f k → 0 as n → ∞.

Let B = { φn : n ∈ I } be an orthonormal basis in the Hilbert space H. I

is the countable index set. Then

(1) hφj |φk i = δjk

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page xiv

xiv Notation
^ X
(2) f= hf |φj iφj
f ∈H j∈I
^ X
(3) hf |gi = hf |φj ihg|φj i
f,g∈H j∈I
 
^
(4)  hf, φj i = 0  ⇒ f = 0
φj ∈B

^ X
(5) kf k2 = |hf, φj i|2 .
f ∈H j∈I

Let f, g ∈ H. Then we have the inequalities

|hf, gi| ≤ kf k · kgk

kf + gk ≤ kf k + kgk

and the equality

kf + gk2 + kf − gk2 = 2(kf k2 + kgk2 ).

We will also use the so-called Dirac notation. Let H be a Hilbert space and
H∗ be the dual space endowed with a multiplication law of the form

(c, φ) = c̄φ

where c ∈ C and φ ∈ H. The inner product can be viewed as a bilinear

form (duality)
h ·| · i : H∗ × H → C
such that the linear maps

hφ| : ψ → hφ|ψi, h · | : H∗ → H 0

|ψi : φ → hφ|ψi, | · i : H → H∗0

where prime denotes the space of linear continuous functionals on the cor-
responding space, are monomorphisms. The vectors hφ| and |ψi are called
bra and ket vectors, respectively. The ket vector |φi is uniquely determined
by a vector φ ∈ H, therefore we can write |φi ∈ H.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 1

Part I

Finite-Dimensional
Hilbert Spaces

1
August 1, 2014 10:17 9in x 6in Scanning Transmission Electron Microscopy. . . b1683-fm

This page intentionally left blank

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 3

Chapter 1

Qubits

1.1 Introduction
A single qubit is a two-state system, such as a two-level atom. The states
(kets) |hi and |vi of the horizontal and vertical polarization of a photon can
also be considered as a two-state system. Another example is the relative
phase and intensity of a single photon in two arms of an interferometer.
The underlying Hilbert space for the qubit is C2 . An arbitrary orthonormal
basis for C2 is denoted by { |0i, |1i }, where (scalar product)

h0|0i = h1|1i = 1, h0|1i = h1|0i = 0.

Any pure quantum state |ψi (qubit) of this system can be written, up to a
phase, as a superposition (linear combination of the states)

|ψi = α|0i + β|1i , |α|2 + |β|2 = 1 , α, β ∈ C.

The classical boolean states, 0 and 1, can be represented by a fixed pair of

orthonormal states of the qubit. The standard basis in C2 is given by
   
1 0
|0i = , |1i =
0 1

and the Hadamard basis in C2 is given by

   
1 1 1 1
|0i = √ , |1i = √ .
2 1 2 −1

3
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 4

4 Problems and Solutions

Up to an overall phase an arbitrary normalized state in C2 can be written

as  iφ 
e cos(θ)
|ψi = .
sin(θ)

For any orthonormal basis {|0i, |1i} in C2 we have

|0ih0| + |1ih1| = I2

where I2 is the 2 × 2 identity matrix. The 2 × 2 matrices

|0ih0|, |1ih1|

are projection matrices with

|0ih0|1ih1| = 02 .

Furthermore
(|0ih1| + |1ih0|)2 = |1ih1| + |0ih0| = I2 .
Given two normalized states |ψi, |φi in C2 , then 0 ≤ |hψ|φi|2 ≤ 1 provides
a probability. Let |ψi ∈ C2 and normalized. Then

ρ = |ψihψ|

is a density matrix (pure state). We have

ρ2 = |ψihψ|ψihψ| = |ψihψ| = ρ.

If the qubit represents a mixed state one uses a two-dimensional density

matrix ρ for its representation. We therefore express one qubit as

1 1
ρ= (I2 + n · σ) ≡ (I2 + n1 σ1 + n2 σ2 + n3 σ3 )
2 2

where n ∈ R3 ,
n · n ≡ n21 + n22 + n23 ≤ 1
and σ = (σ1 , σ2 , σ3 ) denote the Pauli spin matrices
     
0 1 0 −i 1 0
σ1 = , σ2 = , σ3 = .
1 0 i 0 0 −1

For pure states we have n · n = 1 and ρ = |ψihψ|. The Pauli spin matrices
are hermitian and unitary and admit the eigenvalues +1 and −1.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 5

Qubits 5

1.2 Solved Problems

Problem 1. Any normalized state (qubit) in C2 can be written as
 
α
, α, β ∈ C, |α|2 + |β|2 = 1.
β

Find a parameter representation (i) if the underlying field is the set of real
numbers (ii) if the underlying field is the set of complex numbers.

Solution 1. (i) Using α = cos(θ), β = sin(θ) and the identity cos2 (θ) +
sin2 (θ) ≡ 1 for all θ ∈ R we have
 
cos(θ)
.
sin(θ)

With the map θ → θ + π/2 we can construct the orthonormal basis

     
cos(θ) cos(θ + π/2) − sin(θ)
, = .
sin(θ) sin(θ + π/2) cos(θ)

(ii) We have as a representation

 
eiφ cos(θ)
sin(θ)

where θ, φ ∈ R and eiφ e−iφ = 1.

Problem 2. Consider the normalized states in C2 (θ1 , θ2 ∈ [0, 2π))

   
cos(θ1 ) cos(θ2 )
, .
sin(θ1 ) sin(θ2 )

Find the condition on θ1 and θ2 such that the vector

   
cos(θ1 ) cos(θ2 )
+
sin(θ1 ) sin(θ2 )

is normalized.

Solution 2. From the condition that the vector

 
cos(θ1 ) + cos(θ2 )
sin(θ1 ) + sin(θ2 )

is normalized we obtain (sin(θ1 ) + sin(θ2 ))2 + (cos(θ1 ) + cos(θ2 ))2 = 1. Thus

1 1
sin(θ1 ) sin(θ2 ) + cos(θ1 ) cos(θ2 ) = − ⇒ cos(θ1 − θ2 ) = − .
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 6

6 Problems and Solutions

Therefore, θ1 − θ2 = 2π/3 or θ1 − θ2 = 4π/3.

Problem 3. Let { |0i, |1i } be an orthonormal basis in the Hilbert space

R2 and A := |0ih0| + |1ih1|. Consider the three cases
   
1 0
(i) |0i := , |1i :=
0 1
   
1 1 1 1
(ii) |0i := √ , |1i := √
2 1 2 −1
   
cos(θ) sin(θ)
(iii) |0i := , |1i := .
sin(θ) − cos(θ)
Find the matrix representation of A in these bases.

Solution 3. We find
     
1 0 0 0 1 0
(i) A = + =
0 0 0 1 0 1
     
1 1 1 1 1 −1 1 0
(ii) A = + =
2 1 1 2 −1 1 0 1
sin2 (θ)
 2
  
cos (θ) cos(θ) sin(θ) − cos(θ) sin(θ)
(iii) A = +
cos(θ) sin(θ) sin2 (θ) − cos(θ) sin(θ) cos2 (θ)
 
1 0
= .
0 1
For all three cases A = I2 , where I2 is the 2 × 2 unit matrix. Obviously, the
third case contains the first two as special cases. This is the completeness
relation.

Problem 4. Let { |0i, |1i } be an orthonormal basis in the Hilbert space

C2 . The NOT operation (unitary operator) is defined as
|0i → |1i, |1i → |0i.
(i) Find the unitary operator UN OT which implements the NOT operation
with respect to the basis { |0i, |1i }.
(ii) Consider the standard basis
   
1 0
|0i = , |1i = .
0 1
Find the matrix representation of UN OT for this basis.
(iii) Consider the Hadamard basis
   
1 1 1 1
|0i = √ , |1i = √ .
2 1 2 −1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 7

Qubits 7

Find the matrix representation of UN OT for this basis.

Solution 4. (i) Obviously, UN OT = |0ih1| + |1ih0| since h0|0i = h1|1i = 1

and h0|1i = h1|0i = 0.
(ii) For the standard basis we find
 
0 1
UN OT = .
1 0

(iii) For the Hadamard basis we find

 
1 0
UN OT = .
0 −1

Thus we see that the respective matrix representations for the two bases
are different.

Problem 5. Let |0i, |1i be an orthonormal basis in C2 . The Walsh-

Hadamard transform is a 1-qubit operation, denoted by H, and performs
the linear transform
1 1
|0i → √ (|0i + |1i), |1i → √ (|0i − |1i).
2 2
(i) Find the unitary operator UH which implements H with respect to the
basis { |0i, |1i }.
(ii) Find the inverse of this operator.
(iii) Consider the standard basis
   
1 0
|0i = , |1i =
0 1

in C2 . Find the matrix representation of UH for this basis.

(iv) Consider the Hadamard basis
   
1 1 1 1
|0i = √ , |1i = √
2 1 2 −1

in C2 . Find the matrix representation of UH for this basis.

Solution 5. (i) Obviously,

1 1
UH = √ (|0i + |1i)h0| + √ (|0i − |1i)h1|
2 2
1 1
= √ |0i(h0| + h1|) + √ |1i(h0| − h1|).
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 8

8 Problems and Solutions

−1 ∗
(ii) The operator UH is unitary and the inverse is given by UH = UH = UH ,
∗
where denotes the adjoint.
(iii) For the standard basis we find
 
1 1 1
UH = √ .
2 1 −1

(iv) For the Hadamard basis we find

 
1 1 1
UH =√ .
2 1 −1

We see that the matrix representations for each of the two bases are the
same.

Problem 6. The Hadamard operator on one qubit can be written as

1
UH = √ ((|0i + |1i)h0| + (|0i − |1i)h1|).
2

Calculate the states UH |0i and UH |1i. Calculate UH UH .

Solution 6. We obtain the normalized states

1 1
UH |0i = √ (|0i + |1i), UH |1i = √ (|0i − |1i).
2 2

Since h0|0i = h1|1i = 1 and h0|1i = h1|0i = 0 we obtain

UH UH = |0ih0| + |1ih1| = I

where I is the identity operator (2 × 2 unit matrix).

Problem 7. Let σ1 , σ2 , σ3 be the Pauli spin matrices and I2 be the 2 × 2

unit matrix. Consider the Hilbert space C2 and the linear operator (2 × 2
matrix)  
3
1 X
Π(n) := I2 + nj σj 
2 j=1

where n := (n1 , n2 , n3 ) (nj ∈ R) is a unit vector, i.e. n21 + n22 + n23 = 1.

(i) Describe the properties of Π(n), i.e. find Π† (n), tr(Π(n)) and Π2 (n).
(ii) Find the vector (φ, θ ∈ R)
 iφ 
e cos(θ)
Π(n) .
sin(θ)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 9

Qubits 9

Solution 7. (i) For the Pauli matrices we have σ1† = σ1 , σ2† = σ2 , σ3† = σ3 .
Thus Π(n) = Π† (n). Since tr(σ1 ) = tr(σ2 ) = tr(σ3 ) = 0, tr(I2 ) = 2, and the
trace operation is linear, we obtain tr(Π(n)) = 1. Since σ12 = σ22 = σ32 = I2
and
[σ1 , σ2 ]+ = 02 , [σ2 , σ3 ]+ = 02 , [σ3 , σ1 ]+ = 02
where [A, B]+ := AB + BA denotes the anticommutator, the expression
 2
3 3 3 3
1 X 1 1X 1 XX
Π2 (n) = I2 + nj σj  = I2 + nj σj + nj nk σj σk
4 j=1
4 2 j=1 4 j=1
k=1

simplifies to
3 3
1 1X 1X 2
Π2 (n) = I2 + nj σj + n I2 .
4 2 j=1 4 j=1 j

Using n21 + n22 + n23 = 1 we obtain Π2 (n) = Π(n). Thus Π(n) is a projection
matrix.
(ii) We find
 iφ   
e cos(θ) 1 (1 + n3 )eiφ cos(θ) + (n1 − in2 ) sin(θ)
Π(n) = .
sin(θ) 2 (n1 + in2 )eiφ cos(θ) + (1 − n3 ) sin(θ)

Problem 8. The qubit trine is defined by the following states

√ √
1 3 1 3
|ψ0 i = |0i, |ψ1 i = − |0i − |1i, |ψ2 i = − |0i + |1i
2 2 2 2
where { |0i, |1i } is an orthonormal basis. Find the probabilities

|hψ0 |ψ1 i|2 , |hψ1 |ψ2 i|2 , |hψ2 |ψ0 i|2 .

Solution 8. Using h0|0i = 1, h1|1i = 1 and h0|1i = 0 we find

1 1 1
|hψ0 |ψ1 i|2 = , |hψ1 |ψ2 i|2 = , |hψ2 |ψ0 i|2 = .
4 4 4

Problem 9. The kets |hi and |vi are states of horizontal and vertical
polarization, respectively. Consider the normalized states
1 √
|ψ1 i = − (|hi + 3|vi)
2
1 √
|ψ2 i = − (|hi − 3|vi)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 10

10 Problems and Solutions

|ψ3 i = |hi
1 √
|φ1 i = √ (−|hi + 2e−2πi/3 |vi)
3
1 √
|φ2 i = √ (−|hi + 2e+2πi/3 |vi)
3
1 √
|φ3 i = √ (−|hi + 2|vi).
3
Give an interpretation of these states.

Solution 9. Since hh|hi = hv|vi = 1 and hv|hi = hh|vi = 0 we find

1 1 1
hψ1 |ψ2 i = − , hψ1 |ψ3 i = − , hψ2 |ψ3 i = − .
2 2 2
Since the solution to cos(α) = −1/2 is given by α = 120o (2π/3) or α = 240o
(4π/3) we find that the first three states |ψ1 i, |ψ2 i, |ψ3 i correspond to states
of linear polarization separated by 120o (2π/3). We find the scalar product
i
hφ1 |φ2 i = − √ .
3
The states |φ1 i and |φ2 i correspond to elliptic polarization and the third
state |φ3 i corresponds to linear polarization.

Problem 10. Let  

eiφ cos(θ)
|ψi =
sin(θ)
be a normalized state in the Hilbert space C2 , where φ, θ ∈ R. Find the
density matrix ρ := |ψihψ|, tr(ρ) and ρ2 .

Solution 10. Since

hψ| = (e−iφ cos(θ), sin(θ))

we obtain the 2 × 2 density matrix

 
cos2 (θ) eiφ sin(θ) cos(θ)
ρ = |ψihψ| = .
e−iφ sin(θ) cos(θ) sin2 (θ)

Since cos2 (θ) + sin2 (θ) = 1 we obtain that tr(ρ) = 1. With hψ|ψi = 1 we
obtain ρ2 = (|ψihψ|)2 = |ψihψ|ψihψ| = |ψihψ| = ρ. Thus we have a pure
state.

Problem 11. Given the Hamilton operator Ĥ = ~ωσ1 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 11

Qubits 11

(i) Find the solution

|ψ(t)i = e−iĤt/~ |ψ(t = 0)i

of the Schrödinger equation

d
i~ |ψi = Ĥ|ψi
dt
with the initial conditions
 
1
|ψ(t = 0)i = .
0

(ii) Find the probability |hψ(t = 0)|ψ(t)i|2 .

(iii) The solution of the Heisenberg equation of motion

dσ3
i~ = [σ3 , Ĥ](t)
dt
is given by
σ3 (t) = eiĤt/~ σ3 e−iĤt/~
where σ3 (t = 0) = σ3 . Calculate σ3 (t).
(iv) Show that hψ(t = 0)|σ3 (t)|ψ(t = 0)i = hψ(t)|σ3 |ψ(t)i.

Solution 11. (i) The solution of the Schrödinger equation is given by

|ψ(t)i = exp(−iĤt/~)|ψ(t = 0)i.

Since σ12 = I2 we find the unitary matrix

 
cos(ωt) −i sin(ωt)
exp(−iĤt/~) ≡ U (t) = .
−i sin(ωt) cos(ωt)

Thus the normalized state at time t is

   
1 cos(ωt)
|ψ(t)i = U (t) = .
0 −i sin(ωt)

(ii) We find the probability |hψ(t = 0)|ψ(t)i|2 = cos2 (ωt).

(iii) Since the commutators are given by

[σ3 , Ĥ] = ~ω[σ3 , σ1 ] = 2i~ωσ2 , [σ2 , Ĥ] = ~ω[σ2 , σ1 ] = −2i~ωσ3

we obtain the linear system of matrix-valued differential equations

dσ3 dσ2
= 2ωσ2 (t), = −2ωσ3 (t)
dt dt
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 12

12 Problems and Solutions

with the initial conditions σ3 (t = 0) = σ3 and σ2 (t = 0) = σ2 . Here we used

the Heisenberg equation of motion for σ2 to obtain the second differential
equation. The solution of this system of matrix-valued linear differential
equations is given by

σ3 (t) = σ3 cos(2ωt) + σ2 sin(2ωt)

σ2 (t) = σ2 cos(2ωt) − σ3 sin(2ωt).

(iv) We find
hψ(t = 0)|σ3 (t)|ψ(t = 0)i = cos(2ωt)
and
hψ(t)|σ3 |ψ(t)i = cos2 (ωt) − sin2 (ωt) ≡ cos(2ωt).

Problem 12. Consider a Mach-Zehnder interferometer in which the

beam pair spans a two-dimensional Hilbert space with orthonormal basis
{ |0i, |1i }. The state vectors |0i and |1i can be considered as orthonormal
wave packets that move in two given directions defined by the geometry of
the interferometer. We may represent mirrors, beam splitters and relative
UP phase shifts by the unitary matrices
     iχ 
0 1 1 1 1 e 0
UM = , UB = √ , UP =
1 0 2 1 −1 0 1

respectively. Consider the density matrix

ρin = |0ih0|

where { |0i, |1i } denotes the standard basis. Using this basis find

ρout = UB UM UP UB ρin UB† UP† UM

†
UB† .

Give an interpretation of the result.

Solution 12. Since

   
1 1 0
ρin = |0ih0| = (1 0) =
0 0 0

and  
1 eiχ + 1 eiχ − 1
UB UM UP UB =
2 −eiχ + 1 −eiχ − 1
we obtain  
1 1 + cos(χ) i sin(χ)
ρout = .
2 −i sin(χ) 1 − cos(χ)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 13

Qubits 13

This yields the intensity along |0i as I ∝ 1 + cos(χ). Thus the relative UP
phase χ could be observed in the output signal of the interferometer.

Problem 13. Let { |0i, |1i } be an orthonormal basis in C2 .

(i) Find the commutator [ |0ih1|, |1ih0| ].
(ii) Find the operators exp(t|0ih1|) and exp(t|1ih0|).
(iii) Find the operator exp(t|0ih1|) exp(t|1ih0|).
(iv) Find the operator exp(t(|0ih1| + |1ih0|)).
(v) Is exp(t(|0ih1| + |1ih0|)) = exp(t|0ih1|) exp(t|1ih0|)?

Solution 13. (i) We have

h i
|0ih1| , |1ih0| = |0ih0| − |1ih1|

since h0|0i = h1|1i = 1 and h0|1i = h1|0i = 0. We see that the commutator
is nonzero.
(ii) Since h0|1i = h1|0i = 0 we find
∞ j
X t
exp(t|0ih1|) = (|0ih1|)j = I2 + t|0ih1|.
j=0
j!

Analogously
∞ j
X t
exp(t|1ih0|) = (|1ih0|)j = I2 + t|1ih0|.
j=0
j!

(iii) Multiplying the results found above we obtain

exp(t|0ih1|) exp(t|1ih0|) = I2 + t(|0ih1| + |1ih0|) + t2 |0ih0|.

(iv) Since (|0ih1| + |1ih0|)2 = I2 we obtain

∞ ∞
X t2j X t2j+1
exp(t|0ih1| + t|1ih0|) = I2 + (|0ih1| + |1ih0|)
j=0
(2j)! j=0
(2j + 1)!
= cosh(t)I2 + sinh(t)(|0ih1| + |1ih0|).

(v) Clearly when t 6= 0 we have

exp(t(|0ih1| + |1ih0|)) 6= exp(t|0ih1|) exp(t|1ih0|).

Problem 14. Consider the unitary matrix for the NOT gate
 
0 1
UN OT = .
1 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 14

14 Problems and Solutions

Show that we can find a unitary matrix V such that V 2 = UN OT . Thus V

would be the square root NOT gate. What are the eigenvalues of V ?

Solution 14. We find the unitary matrix

 
1 1+i 1−i
V = .
2 1−i 1+i

Obviously −V is also a square root. The eigenvalues of V are 1 and i. The

−V are −1√and −i. Note that the eigenvalues of UN OT are
eigenvalues of √
1 and −1 and 1 = ±1, −1 = ±i.

Problem 15. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Let n be a unit

vector in R3 . We define the operator

Σ := n · σ ≡ n1 σ1 + n2 σ2 + n3 σ3 .

(i) Calculate the matrix Σ2 . From this result and the fact that Σ is hermi-
tian show that Σ is unitary. Find the eigenvalues of Σ.
(ii) Let  
1
|ψi = .
0
Calculate the state Σ|ψi and the probability |hψ|Σ|ψi|2 .

Solution 15. (i) Using n21 + n22 + n23 = 1, σ12 = σ22 = σ32 = I2 and

σ1 σ2 + σ2 σ1 = 02 , σ1 σ3 + σ3 σ1 = 02 , σ2 σ3 + σ3 σ2 = 02

we obtain

Σ2 = (n1 σ1 + n2 σ2 + n3 σ3 )2
= (n21 + n22 + n23 )I2
+n1 n2 (σ1 σ2 + σ2 σ1 ) + n1 n3 (σ3 σ1 + σ1 σ3 ) + n2 n3 (σ2 σ3 + σ3 σ2 )
= I2 .

Since Σ is hermitian, i.e. Σ = Σ∗ and Σ2 = I2 we find that Σ is a unitary

matrix with Σ = Σ−1 . Since Σ is hermitian and unitary the eigenvalues
λ1 , λ2 can only be ±1. Since tr(Σ) = 0 = λ1 + λ2 we obtain that the
eigenvalues are +1 and −1.
(ii) We find      
0 0 1
Σ|ψi = n1 + n2 + n3 .
1 i 0
It follows that |hψ|Σ|ψi|2 = n23 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 15

Qubits 15

Problem 16. Let n be a unit vector in R3 , σ = (σ1 , σ2 , σ3 ) and

n · σ := n1 σ1 + n2 σ2 + n3 σ3 .

(i) Find the unitary matrix exp(iθn · σ), where θ ∈ R.

(ii) Find the normalized state
 
1
exp(iθn · σ) .
0

Solution 16. (i) Since

3
X
σj σk = δjk I2 + i jk` σ`
`=1

where 123 = 231 = 312 = 1, 321 = 213 = 132 = −1 and 0 otherwise, we

obtain

exp(iθn · σ) = I2 cos(θ) + i(n · σ) sin(θ)

 
cos(θ) + in3 sin(θ) i(n1 − in2 ) sin(θ)
= .
i(n1 + in2 ) sin(θ) cos(θ) − in3 sin(θ)

Note that we could also use (n · σ)2 = I2 to find the result.

(ii) Using (i) we find the state
   
1 cos(θ) + in3 sin(θ)
exp(iθn · σ) = .
0 i(n1 + in2 ) sin(θ)

Problem 17. Consider the Hamilton operator

 
0 α
Ĥ = ~ω
α 1

where α ≥ 0. Find α where the energy gap between the two energy levels
is the smallest.

Solution 17. From the eigenvalue equation we find E 2 − ~ωE = ~2 ω 2 α2 .

Consequently
p p
E0 (α) = ~ω(1 − 1 + α2 ), E1 (α) = ~ω(1 + 1 + α2 ).

Thus p
E1 (α) − E0 (α) = 2~ω 1 + α2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 16

16 Problems and Solutions

Therefore the shortest energy gap is for α = 0.

Problem 18. Consider the Hamilton operator

 
~ω ∆
Ĥ = ~ωσ3 + ∆σ1 =
∆ −~ω

where ∆ ≥ 0.
(i) Find the eigenvalues and the normalized eigenvectors of Ĥ.
(ii) Use the Cayley-Hamilton theorem to calculate exp(−iĤt/~).

Solution 18. (i) From det(Ĥ − EI2 ) = 0 we find the two eigenvalues
p
E± = ± ~2 ω 2 + ∆2 .
√
We set E := ~2 ω 2 + ∆2 . Then from the eigenvalue equation
    
~ω ∆ u1 u1
= E+
∆ −~ω u2 u2

for the eigenvalue E+ = E we find ∆u2 = (E −~ω)u1 . Thus the eigenvector

is given by  
∆
.
E − ~ω
After normalization we have
 
1 ∆
.
E − ~ω
p
∆2 + (E − ~ω)2

Analogously we find for the eigenvalue E− = −E the normalized eigenvec-

tor  
1 ∆
.
∆2 + (E + ~ω)2 −E − ~ω
p

(ii) Since E+ 6= E− and E+ = E, E− = −E we have to solve the system of

equations
e−iEt/~ = c0 + c1 E, eiEt/~ = c0 − c1 E
for c0 and c1 . Then
 
−iĤt/~ c0 + c1 ~ω c1 ∆
e = c0 I2 + c1 Ĥ = .
c1 ∆ c0 − c1 ~ω

The solution of the system of equations is given by

e−iEt/~ − eiEt/~ −i sin(Et/~)

c0 = cos(Et/~), c1 = = .
2E E
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 17

Qubits 17

Thus
 
cos(Et/~) − i sin(Et/~)~ω/E −i sin(Et/~)∆/E
e−iĤt/~ = .
−i sin(Et/~)∆/E cos(Et/~) + i sin(Et/~)~ω/E

Obviously, exp(−iĤt/~) is a unitary matrix.

Problem 19. Consider the Pauli spin matrices σ1 , σ2 , σ3 . Can one find
an α ∈ R such that exp(iασ3 )σ1 exp(−iασ3 ) = σ2 ?

Solution 19. We have

 
0 e2iα
exp(iασ3 )σ1 exp(−iασ3 ) = .
e−2iα 0
Thus we have to solve the equations exp(2iα) = −i, exp(−2iα) = i. For
α ∈ [0, 2π) we obtain α = 3π/4.

Problem 20. Let n and m be a unit vectors in R3 , σ = (σ1 , σ2 , σ3 ) and

n · σ := n1 σ1 + n2 σ2 + n3 σ3 .
Calculate the commutator [n · σ, m · σ].

Solution 20. We find

[n · σ, m · σ] = 2i((n2 m3 − m2 n3 )σ1 + (n3 m1 − m3 n1 )σ2
+(n1 m2 − m1 n2 )σ3 )
= 2i(n × m) · σ
where × denotes the vector product. The vector n × m is perpendicular to
the plane spanned by the vectors n and m.

Problem 21. Let |ψ1 i and |ψ2 i be two normalized states in a Hilbert
space H. A distance d with 0 ≤ d ≤ π/2 can be defined as
cos2 (d) := |hψ1 |ψ2 i|2 .
Let H = C2 and consider the normalized states
   
1 1 1 1
|ψ1 i = √ , |ψ2 i = √ .
2 1 2 −1
Find d.

Solution 21. Since hψ1 |ψ2 i = 0 we have cos2 (d) = 0 and therefore
d = π/2.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 18

18 Problems and Solutions

Problem 22. Let ρ1 and ρ2 be density matrices in the same Hilbert

space. The Bures distance between the two density matrices is defined as
q
1/2 1/2
DB (ρ1 , ρ2 ) := 2(1 − tr((ρ1 ρ2 ρ1 )1/2 )).

Consider the density matrices

   
1 0 1/2 0
ρ1 = , ρ2 =
0 0 0 1/2

acting in the Hilbert space C2 . Find the Bures distance.

Solution 22. Since  

1/2 1 0
ρ1 = ρ1 =
0 0
we obtain  
1/2 1/2 1/2 0
ρ1 ρ2 ρ1 = .
0 0
q √
Thus DB (ρ1 , ρ2 ) = 2(1 − 1/ 2).

Problem 23. (i) Consider the Hilbert space C2 . Show that

   
1 1 −i 1 1 i
ΠS = , ΠA =
2 i 1 2 −i 1
are projection matrices.
(ii) Decompose the Hilbert space into sub-Hilbert spaces using the result
from (i).

Solution 23. (i) We have

ΠS = Π∗S , Π2S = ΠS , ΠA = Π∗A , Π2A = ΠA

and ΠS + ΠA = I2 , ΠS ΠA = 02 .
(ii) Consider the normalized state
 iφ 
e sin(θ)
|ψi = .
cos(θ)

Then
   
eiφ sin(θ) − i cos(θ) eiφ sin(θ) + i cos(θ)
ΠS |ψi = , ΠA |ψi =
ieiφ sin(θ) + cos(θ) −ieiφ sin(θ) + cos(θ)

with hψ|ΠA ΠS |ψi = 0.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 19

Qubits 19

Problem 24. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Consider the

normalized state in C2
 iφ 
e cos(θ)
ψ= ⇒ ψ ∗ = ( e−iφ cos(θ) sin(θ) ) .
sin(θ)

Find the vector v = (v1 v2 v3 )T in R3 with

v1 = ψ ∗ σ1 ψ, v2 = ψ ∗ σ2 ψ, v3 = ψ ∗ σ3 ψ.

Is the vector v normalized?

Solution 24. Utilizing that

eiφ + e−iφ ≡ 2 cos(φ), ieiφ − ie−iφ ≡ −2 sin(φ),

1
cos(θ) sin(θ) ≡ sin(2θ), cos2 (θ) − sin2 (θ) ≡ cos(2θ)
2
we obtain

v1 = ψ ∗ σ1 ψ = cos(φ) sin(2θ)
v2 = ψ ∗ σ2 ψ = − sin(φ) sin(2θ)
v3 = ψ ∗ σ3 ψ = cos(2θ).

The vector    
v1 cos(φ) sin(2θ)
v =  v2  =  − sin(φ) sin(2θ) 
v3 cos(2θ)
is normalized, i.e. kvk2 = 1.

Problem 25. (i) Consider the symmetric matrix over R

 
h11 h12
H=
h12 h22

and the normalized state

 
cos(θ)
|ψi = .
sin(θ)

Calculate the variance VH (|ψi) := hψ|H 2 |ψi − (hψ|H|ψi)2 .

(ii) Consider the Hadamard matrix
 
1 1 1
H=√
2 1 −1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 20

20 Problems and Solutions

and the normalized state

 
cos(θ)
|ψi = .
sin(θ)

Calculate the variance VH (ψ) := hψ|H 2 |ψi − (hψ|H|ψi)2 and discuss the
dependence on θ.

Solution 25. (i) We find

hψ|H 2 |ψi = h211 cos2 (θ) + h222 sin2 (θ) + h212 + 2h12 (h11 + h22 ) cos(θ) sin(θ)

and

(hψ|H|ψi)2 = h211 cos4 (θ) + h222 sin4 (θ) + 2h11 h22 cos2 (θ) sin2 (θ)
+4h11 h12 cos3 (θ) sin(θ) + 4h22 h12 cos(θ) sin3 (θ).

Thus

VH (|ψi) = (h211 + h222 − 2h11 h22 ) sin2 (θ) cos2 (θ) + h212 (1 − 4 cos2 (θ) sin2 (θ))
+2h12 h11 sin(θ) cos(θ)(1 − 2 cos2 (θ))
+2h12 h22 sin(θ) cos(θ)(1 − 2 sin2 (θ)).

(ii) Using that H 2 = I2 we have

1
VH (ψ) = hψ|I2 |ψi − (hψ|H|ψi)2 = 1 − (cos(2θ) + sin(2θ))2
2
1
= (1 − sin(4θ)).
2
For θ = 0 we have VH (ψ) = 1/2. The minimum value is 0, for example for
θ = π/8. The maximum value is 1, for example for θ = 3π/8.

Problem 26. Let c† , c be Fermi creation and annihilation operators with

[c† , c]+ = c† c + cc† = I

and (c† ) = 0, c2 = 0, where I is the identity operator and 0 the zero

operator. Consider the Hamilton operator

Ĥ = ~ω1 (eiφ c† + e−iφ c) + ~ω2 c† c

and the basis |0i, c† |0i. The dual basis is h0|, h0|c. Find the matrix repre-
sentation of Ĥ and the eigenvalues of the matrix.

Solution 26. We obtain

Ĥ|0i = ~ω1 eiφ c† |0i

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 21

Qubits 21

Ĥc† |0i = ~ω1 e−iφ |0i + ~ω2 c† |0i.

Hence we obtain the matrix representation of Ĥ

~ω1 e−iφ
 
0
.
~ω1 eiφ ~ω2

The eigenvalues are given by

1
q
~ω2
E± = ± 4~2 ω12 + ~2 ω22 .
2 2

Problem 27. Let σ1 , σ2 , σ3 be the Pauli spin matrices and I2 the 2 × 2

identity matrix. Find the eigenvalues and normalized eigenvectors of the
Hamilton operator

Ĥ = ε0 I2 + ~ωσ3 + ∆1 σ1 + ∆2 σ2

where ε0 > 0. Are the normalized eigenvectors orthonormal to each other?

Solution 27. In matrix form we have the (hermitian) 2 × 2 matrix

 
ε0 + ~ω ∆1 − i∆2
Ĥ = .
∆1 + i∆2 ε0 − ~ω

From

det(Ĥ − EI2 ) = (ε + ~ω − E)(ε − ~ω − E) − (∆1 + i∆2 )(∆1 − i∆2 ) = 0

we obtain the characteristic equation

E 2 − 2ε0 E + ε20 = ~2 ω 2 + ∆21 + ∆22 .

Thus the two eigenvalues E+ , E− are

q
E± = ε0 ± ~2 ω 2 + ∆21 + ∆22 .
p
Let S := ~2 ω 2 + ∆21 + ∆22 . For the eigenvector of E+ we have to solve
    
ε0 + ~ω ∆1 − i∆2 v1 v1
= E+
∆1 + i∆2 ε0 − ~ω v2 v2
or

(ε0 + ~ω)v1 + (∆1 − i∆2 )v2 = E+ v1 = (ε0 + S)v1

(∆1 + i∆2 )v1 + (ε0 − ~ω)v2 = (ε0 + S)v2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 22

22 Problems and Solutions

We can set v1 = S + ~ω. Thus v2 = ∆1 + i∆2 . After normalizing we have

the eigenvector
 
1 S + ~ω
.
(S + ~ω)2 + ∆2 + ∆2 ∆1 + i∆2
p
1 2

Analogously for E− we obtain the normalized eigenvector

 
1 ∆1 − i∆2
.
(S + ~ω)2 + ∆21 + ∆22 −S − ~ω
p

Obviously the two eigenvectors are orthonormal to each other, i.e. the
scalar product vanishes. The eigenvectors do not depend on ε0 .

Problem 28. Let  and B̂ be n × n hermitian matrices. Let |ψi be a

normalized state in the Hilbert space Cn . Then we have the inequality

1
(∆Â)(∆B̂) ≥ |h[Â, B̂]i|
2
where q q
∆ := hÂ2 i − hÂi2 , ∆B̂ := hB̂ 2 i − hB̂i2
and
hÂi := hψ|Â|ψi, hB̂i := hψ|B̂|ψi.
Consider the hermitian spin- 21 matrices
     
1 0 1 1 0 −i 1 1 0
S1 = , S2 = , S3 = .
2 1 0 2 i 0 2 0 −1

Let  = S1 and B̂ = S2 . Find states |ψi such that

1
(∆Â)(∆B̂) = |h[Â, B̂]i|
2
i.e. the inequality given above should be an equality.

Solution 28. For the commutator we find [S1 , S2 ] = iS3 . Now

1
S12 = S22 = S32 = I2 .
4
We set  
c1
|ψi = , hψ| = ( c∗1 c∗2 )
c2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 23

Qubits 23

with c1 c∗1 + c2 c∗2 = 1 (normalization). Thus we have for the right-hand side
of the equality
    
1 1 1 1 0 c1  1 2
|h[Â, B̂]i| = |hψ|iS3 |ψi = ( c∗1 c∗2 ) = |r − r22 |
2 2 4 0 −1 c2  4 1

where we set c1 = r1 eiφ1 , c2 = r2 eiφ2 . Now we have

r
1 1
q
∆S1 = h I2 i − hS1 i2 = 1 − (c∗1 c2 + c1 c∗2 )2
4 2
r
1 1
q
∆S2 = h I2 i − hS2 i2 = 1 − (c1 c∗2 − c∗1 c2 )2 .
4 2
Thus for the left-hand side we find
1 1
q q
(∆S1 )(∆S2 ) = √ 1 − (c1 c∗2 )2 − (c∗1 c2 )2 = √ 1 − 2r12 r22 cos(2(φ1 − φ2 )).
2 2 2 2
Thus the condition from the equality is

1 1
q
√ 1 − 2r12 r22 cos(2(φ1 − φ2 )) = |r12 − r22 |.
2 2 4

Problem 29. Consider a d-dimensional Hilbert space with two orthonor-

mal bases
|b11 i, |b12 i, . . . |b1d i ∈ B1

|b21 i, |b22 i, ... |b2d i ∈ B2 .

The two bases are said to be mutually unbiased bases if

1
|hb2j |b1k i| = √
d

for all j, k = 1, . . . , d and h | i denotes the scalar product in the Hilbert

space. Consider the Hilbert space M2 (C) of 2 × 2 matrices over C, where
the scalar product is defined as

hA|Bi = tr(AB ∗ ), A, B ∈ M2 (C).

Thus d = dim(M2 (C)) = 4. The standard basis in this Hilbert space is

given by
       
1 0 0 1 0 0 0 0
E11 = , E12 = , E21 = , E22 = .
0 0 0 0 1 0 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 24

24 Problems and Solutions

Let UH be the Hadamard matrix

 
1 1 1 ∗
UH = √ , UH = UH .
2 1 −1
Show that the matrices E
ejk (j, k = 1, 2)

ejk = UH Ejk U ∗ ,
E j, k = 1, 2
H

and the standard basis form mutually unbiased bases.

Solution 29. Straightforward calculations yield

 
e11 = UH E11 U ∗ = 1 1 1
E H
2 1 1
 
∗ 1 1 −1
E
e12 = UH E12 UH =
2 1 −1
 
e21 = UH E21 U ∗ = 1
E
1 1
H
2 −1 −1
 
∗ 1 1 −1
E
e22 = UH E22 UH = .
2 −1 1
It follows that
|tr(E1j Ee2k )| = 1 for all j, k = 1, 2.
2
Thus we have mutually unbiased bases. Apply the vec-operator to the
matrices Ejk and Eejk (j, k = 1, 2) to find mutually unbiased bases in the
4
Hilbert space C .

Problem 30. Let x0 = ct. We define a linear bijection, h, between R4

and H(2), the set of complex 2 × 2 hermitian matrices, by
 
x0 + x1 x2 − ix3
(x0 , x1 , x2 , x3 ) → .
x2 + ix3 x0 − x1
We denote the matrix on the right hand side by H.
(i) Show that the matrix can be written as a linear combination of the Pauli
spin matrices and the identity matrix I2 .
(ii) Find the inverse map.
(iii) Calculate the determinant of the 2 × 2 hermitian matrix H. Discuss.

Solution 30. (i) We have H = x0 I2 + x1 σ3 + x2 σ1 + x3 σ2 .

(ii) Consider (a, b ∈ R)
   
a c x0 + x1 x2 − ix3
= .
c∗ b x2 + ix3 x0 − x1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 25

Qubits 25

Comparing the entries we obtain

a+b a−b c + c∗ c∗ − c
x0 = , x1 = , x2 = , x3 = .
2 2 2 2i
(iii) We obtain
det(H) = x20 − x21 − x22 − x23 .
This is the Lorentz metric. Let U be a unitary 2 × 2 matrix. Then
det(U HU ∗ ) = det(H).

Programming Problems

Problem 1. Consider the unary gates (2 × 2 unitary matrices)

   
0 1 1 1 1
N= , H=√ ,
1 0 2 1 −1
   
1 0 1 0
V = , W =
0 eiπ/2 0 eiπ/4
and the normalized state
 
1 1
|ψi = √ .
2 1
Calculate the state N HV W |ψi and the expectation value hψ|N HV W |ψi.

Solution 1. Applying the Maxima program

/* unary.mac */
N: matrix([0,1],[1,0]);
H: matrix([1/sqrt(2),1/sqrt(2)],[1/sqrt(2),-1/\sqrt(2)]);
V: matrix([1,0],[0,exp(%i*%pi/2)]);
W: matrix([1,0],[0,exp(%i*%pi/4)]);
psi: matrix([1/sqrt(2)],[1/sqrt(2)]);
psiT: matrix([1/sqrt(2),1/sqrt(2)]);
R1: N . H . V . W; R1: ratsimp(R1);
R2: R1 . psi; R2: ratsimp(R2);
R3: psiT . R2; R3: ratsimp(R3);
we find the unitary matrix
 
1 1 −ei3π/4
N HV W = √
2 1 ei3π/4
the normalized state
 
1 1 − ei3π/4
N HV W |ψi =
2 1 + ei3π/4
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 26

26 Problems and Solutions

and
1
hψ|N HV W |ψi = √ .
2

Problem 2. Consider the Pauli spin matrix σ1 and the normalized state
in C2  
cos(θ)
|ψi = .
sin(θ)
Calculate the variance
Vσ1 (ψ) := hψ|σ12 |ψi − (hψ|σ1 |ψi)2
and discuss the dependence on θ

Solution 2. The following Maxima program

/* Variancesig1 */
sig1: matrix([0,1],[1,0]);
sig2: matrix([0,-%i],[%i,0]);
sig3: matrix([1,0],[0,-1]);
psi: matrix([cos(theta)],[sin(theta)]);
psiT: transpose(psi);
Vs1: psiT . (sig1 . sig1) . psi - (psiT . sig1 . psi)^2;
Vs1: trigsimp(Vs1);
D1: diff(Vs1,theta);
list: solve(D1=0,theta);
theta1: rhs(part(list,1));
theta2: rhs(part(list,2));
theta3: rhs(part(list,3));
theta4: rhs(part(list,4));
r11: subst(theta1,theta,Vs1); r12: subst(theta2,theta,Vs1);
r13: subst(theta3,theta,Vs1); r14: subst(theta4,theta,Vs1);
provides
Vσ1 (ψ) := hψ|σ12 |ψi − (hψ|σ1 |ψi)2 = 1 + 4(cos4 (θ) − cos2 (θ)).
Differentiation with respect to θ and solving the resulting equation provides
that the variance is 1 for θ = 0 and θ = π/2. The variance is 0 for θ = π/4
and θ = 3π/4.

Problem 3. Find the eigenvalues and normalized eigenvectors of the

Hamilton operator
 
Ĥ 1 1 1
K̂ = =√ .
~ω 2 1 −1

Solution 3. The Maxima program

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 27

Qubits 27

/* Hamilton1.mac */
H: matrix([1,1],[1,-1])/sqrt(2);
list: eigenvectors(H);
p1: part(list,1);
p11: part(p1,1);
lam1: part(p11,1);
lam2: part(p11,2);
p2: part(list,2);
v1: part(p2,1); v1: part(v1,1);
v2: part(p2,2); v2: part(v2,1);
v2T: transpose(v2);
scalar: v1 . v2T;
scalar: ratsimp(scalar);
provides the eigenvalues λ1 = −1, λ2 = 1 with the corresponding (nonnor-
malized) eigenvectors
   
√ 1 √ 1
v1 = , v2 = .
− 2−1 2−1
The two eigenvectors are orthogonal to each other, i.e. scalar=0.

1.3 Supplementary Problems

Problem 1. Consider the map f : C2 → R3 defined by

 
  sin(2θ) cos(φ)
cos(θ)
f : 7→  sin(2θ) sin(φ)  .
eiφ sin(θ)
cos(2θ)
Are the vectors in C2 and R3 normalized? Consider the four normalized
vectors in C2
       
1 1 1 1 1 1 1 1
√ , √ , √ , √ .
2 1 2 −1 2 i 2 −i
Find the vectors in R3 .

Problem 2. (i) Let x1 , x2 , x3 ∈ R and σ1 , σ2 , σ3 be the Pauli spin

matrices. Show that
sin(r)
ei(x1 σ1 +x2 σ2 +x3 σ3 ) = cos(r)I2 + i (x1 σ1 + x2 σ2 + x3 σ3 )
 r 
cos(r) + ix3 sin(r)/r i(x1 − ix2 ) sin(r)/r
=
i(x1 + ix2 ) sin(r)/r cos(r) − ix3 sin(r)/r
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 28

28 Problems and Solutions

p
where r := x21 + x22 + x23 .
(ii) Let y1 , y2 , y3 ∈ R and

X := x1 σ1 + x2 σ2 + x3 σ3 , Y := y1 σ1 + y2 σ2 + y3 σ3 .

Consider the maps

  

x1 y1
X ↔ x =  x2  , Y ↔ y =  y2  .
x3 y3

Let x · y := x1 y1 + x2 y2 + x3 y3 (scalar product). Show that

1
x·y = tr(XY ).
2
(iii) Show that
 
x2 y3 − x3 y2
i
− [X, Y ] ↔ x × y =  x3 y1 − x1 y3  .
2
x1 y2 − x2 y1

Problem 3. Let Ĥ be a 2 × 2 hermitian matrix. Consider the normalized

state  iφ 
e cos(θ)
|ψi =
sin(θ)
in the Hilbert space C2 . Assume that

hψ|Ĥ|ψi = ~ω cos(φ) sin(2θ), hψ|Ĥ 2 |ψi = ~2 ω 2 .

Reconstruct the hermitian matrix Ĥ from these three assumptions. Note

that
1
cos(θ) sin(θ) ≡ sin(2θ), eiφ = cos(φ)+i sin(φ), e−iφ = cos(φ)−i sin(φ).
2
Show that H = ~ωσ1 .

Problem 4. Let A, B be n × n matrices over C. Let v be a normalized

(column) vector in Cn . Let hAi := v∗ Av and hBi := v∗ Bv. We have the
identity

AB ≡ (A − hAiIn )(B − hBiIn ) + AhBi + BhAi − hAihBiIn .

We approximate AB as AB ≈ AhBi + BhAi − hAihBiIn .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 29

Qubits 29

(i) Let n = 2 and

 
1 1
A = σ1 , B = σ2 , v= √ .
2 1

Find AB and AhBi + BhAi − hAihBiI2 and the distance (Frobenius norm)
between the two matrices.
(ii) Let n = 2 and
 
1 1
A = σ1 , B = σ2 , v = √ .
2 −1

Find AB and AhBi + BhAi − hAihBiI2 and the distance (Frobenius norm)
between the two matrices.
(iii) Consider the case
 
cos(θ)
A = σ1 , B = σ2 , v = .
sin(θ)

Find AB and AhBi + BhAi − hAihBiI2 and the distance (Frobenius norm)
between the two matrices.

Problem 5. Let α ∈ R. Show that the vectors

   
1 cosh(α) 1 sinh(α)
v1 = q , v2 = q
sinh(α) − cosh(α)
1 + 2 sinh2 (α) 1 + 2 sinh2 (α)

form an orthonormal basis in C2 . Find v1 v1T , v2 v2T , v1 v2T .

Problem 6. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Show that

1 1
Π+ = (I2 + σj ), Π− = (I2 − σj )
2 2
(j = 1, 2, 3) are projection matrices. Find the vectors
   
1 1 1 1
Π+ √ , Π− √ .
2 1 2 1

Are the vectors normalized?

Problem 7. Given two arbitrary normalized states |ψi and |φi in C2 .

Find a 2 × 2 unitary matrix U such that |ψi = U |φi, i.e. U must be
expressed in terms of the components of the states |ψi and φi. Since U is
unitary we have U −1 = U ∗ .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 30

30 Problems and Solutions

Problem 8. Let A, B be 2 × 2 hermitian matrices and

 
cos(θ)
|ψi = .
sin(θ)

Find the minima of the function

f (θ) = kAB − Ahψ|B|ψi − hψ|A|ψiB + hψ|A|ψihψ|B|ψiI2 k

where k.k denotes the norm.

Problem 9. Let |0i, |1i be an orthonormal basis in C2 and z00 , z01 , z10 ,
z11 be complex numbers. Calculate

exp(z00 |0ih0| + z01 |0ih1| + z10 |1ih0| + z11 |1ih1|).

Then set

z00 = −i~ω1 , z01 = −z10 = −i~ω2 , z11 = −i~ω3 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 31

Chapter 2

Kronecker Product and

Tensor Product

2.1 Introduction
Let H1 and H2 be two Hilbert spaces and H be a third Hilbert space defined
in terms of H1 and H2 with the following specifications. For each pair of
vectors f1 , f2 in H1 , H2 , respectively, there are vectors in H denoted by
f1 ⊗ f2 and g1 ⊗ g2 , respectively such that
hf1 ⊗ f2 |g1 ⊗ g2 i = hf1 |g1 iH1 hf2 |g2 iH2
where hf1 |g1 i is the scalar product in the Hilbert space H1 . The vector
f1 ⊗ f2 is called the tensor product of the vectors f1 and f2 . The Hilbert
space H consists of the linear combinations of the vectors f1 ⊗ f2 together
with the strong limits of their Cauchy sequences. We term H the tensor
product of H1 and H2 and denote it by H1 ⊗H2 . Given a basis { |φi i : i ∈ I }
in the Hilbert H1 and a basis { |ψj i : j ∈ J } in the Hilbert space H2 we
can construct a basis
{ |φi i ⊗ |ψj i : i ∈ I, j ∈ J }
in the product Hilbert space. The tensor product is associative and dis-
tributive. If Â1 and Â2 are linear operators in H1 and H2 , respectively, we
define the operator Â1 ⊗ Â2 in H1 ⊗ H2 by the formula
(Â1 ⊗ Â2 )(f1 ⊗ f2 ) = (Â1 f1 ) ⊗ (Â2 f2 ) .

31
January 30, 2018 15:21 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 32

32 Problems and Solutions

Â1 ⊗ Â2 is called the tensor product of Â1 and Â2 . Similarly we can define
the tensor product of n Hilbert spaces.

For the finite-dimensional Hilbert spaces Cn and Rn the tensor product is

realized by the Kronecker product.

Let A be an m×n matrix and B be an r ×s matrix. The Kronecker product

of A and B is defined as the (m · r) × (n · s) matrix

a11 B a12 B ... a1n B

 
 a21 B a22 B ... a2n B 
A ⊗ B := 
 ... .. .. ..  .
. . . 
am1 B am2 B ... amn B

We have the following properties.

1) If vj (j = 1, . . . , n) is an orthonormal basis in Cn , then

vj ⊗ vk , (j, k = 1, . . . , n)
2
is an orthonormal basis in Cn .
2) If A, B are normal matrices, then A ⊗ B is a normal matrix.
3) If A, B are hermitian matrices, then A ⊗ B is a hermitian matrix.
4) If A, B are unitary matrices, then A ⊗ B is a unitary matrix.
5) If A, B are projection matrices, then A ⊗ B is a projection matrix.
6) If A, B are nilpotent matrices, then A ⊗ B is a nilpotent matrix.
7) If P1 , P2 are permutation matrices, then P1 ⊗P2 is a permutation matrix.
8) If A and B are invertible, then A ⊗ B is invertible with

(A ⊗ B)−1 ≡ A−1 ⊗ B −1 .

Let A, B, C, D be matrices and assume that the matrix products AC and

BD exist. Then

(A ⊗ B)(C ⊗ B) = (AC) ⊗ (BD).

Let A be an m × m matrix and B be an n × n matrix. The underlying field

is C. Let Im , In be the m × m and n × n unit matrix, respectively. Then

tr(A ⊗ B) = tr(A)tr(B)

and
tr(A ⊗ In + Im ⊗ B) = n tr(A) + m tr(B).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 33

Kronecker Product and Tensor Product 33

2.2 Solved Problems

Problem 1. (i) Let
   
1 0
|φ1 i := , |φ2 i := .
0 1

Thus the set { |φ1 i, |φ2 i } forms a basis in C2 (the standard basis). Calcu-
late the vectors in C4

|φ1 i ⊗ |φ1 i, |φ1 i ⊗ |φ2 i, |φ2 i ⊗ |φ1 i, |φ2 i ⊗ |φ2 i

and interpret the result.

(ii) Consider the Pauli matrices σ1 and σ3 . Find σ1 ⊗ σ3 and σ3 ⊗ σ1 and
discuss. Both σ1 and σ3 are hermitian. Are σ1 ⊗ σ3 and σ3 ⊗ σ1 hermitian?
Both σ1 and σ3 are unitary. Is σ1 ⊗ σ3 and σ3 ⊗ σ1 unitary?

Solution 1. (i) We obtain

1 0
   
0 1
|φ1 i ⊗ |φ1 i =   , |φ1 i ⊗ |φ2 i =   ,
0 0
0 0
0 0
   
0 0
|φ2 i ⊗ |φ1 i =   , |φ2 i ⊗ |φ2 i =   .
1 0
0 1
Thus we find the standard basis in C4 from the standard basis in C2 .
(ii) We obtain
0 0 1 0
 
 0 0 0 −1 
σ1 ⊗ σ3 = 
1 0 0 0

0 −1 0 0
and
0 1 0 0
 
1 0 0 0 
σ3 ⊗ σ1 =  .
0 0 0 −1
0 0 −1 0
We note that σ1 ⊗ σ3 6= σ3 ⊗ σ1 . σ1 ⊗ σ3 and σ3 ⊗ σ1 are hermitian. σ1 ⊗ σ3
and σ3 ⊗ σ1 are unitary.

Problem 2. Given the orthonormal basis

 iφ   
e cos(θ) − sin(θ)
|ψ1 i = , |ψ2 i =
sin(θ) e−iφ cos(θ)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 34

34 Problems and Solutions

in the Hilbert space C2 . Use this basis to find a basis in C4 .

Solution 2. A basis in C4 is given by

{ |ψ1 i ⊗ |ψ1 i, |ψ1 i ⊗ |ψ2 i, |ψ2 i ⊗ |ψ1 i, |ψ2 i ⊗ |ψ2 i }

since
(hψj | ⊗ hψk |)(|ψm i ⊗ |ψn i) = δjm δkn
where j, k, m, n = 1, 2.

Problem 3. A system of n-qubits can be represented by a finite-dimensional

Hilbert space over the complex numbers of dimension 2n . A state |ψi of
the system is a superposition of the basic states
1
X
|ψi = cj1 j2 ...jn |j1 i ⊗ |j2 i ⊗ · · · ⊗ |jn i.
j1 ,j2 ,...,jn =0

In a short cut notation this state is written as

1
X
|ψi = cj1 j2 ...jn |j1 j2 . . . jn i.
j1 ,j2 ,...,jn =0

Consider as a special case the state

1 1
|ψi = (|0i⊗|0i+|0i⊗|1i+|1i⊗|0i+|1i⊗|1i) ≡ (|00i+|01i+|10i+|11i)
2 2
in the Hilbert space H = C4 (n = 2). Can this state be written as a product
state?

Solution 3. Yes, the state can be written as product state. We have

1 1
√ (|0i + |1i) ⊗ √ (|0i + |1i).
2 2

Problem 4. The single-bit Walsh-Hadamard transform is the unitary

map W given by
1 1
W |0i = √ (|0i + |1i), W |1i = √ (|0i − |1i).
2 2
The n-bit Walsh-Hadamard transform Wn is defined as

Wn := W ⊗ W ⊗ · · · ⊗ W (n − times).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 35

Kronecker Product and Tensor Product 35

Consider n = 2. Find the normalized state W2 (|0i ⊗ |0i).

Solution 4. We have

W2 (|0i ⊗ |0i) = (W ⊗ W )(|0i ⊗ |0i) = W |0i ⊗ W |0i.

Thus
1
W2 (|0i ⊗ |0i) = ((|0i + |1i) ⊗ (|0i + |1i)).
2
Finally
1
W2 (|0i ⊗ |0i) = (|0i ⊗ |0i + |0i ⊗ |1i + |1i ⊗ |0i + |1i ⊗ |1i).
2
W2 generates a linear combination of all states. This also applies to Wn .

Problem 5. Consider the spin matrix S1 for spin- 21

 
1 1 0 1
S1 = σ1 =
2 2 1 0

with the eigenvalues 1/2 and −1/2 and the corresponding normalized eigen-
vectors    
1 1 1 1
e1/2 = √ , e−1/2 = √ .
2 1 2 −1
Do the four vectors
1 1
√ (e1/2 ⊗ e1/2 + e−1/2 ⊗ e−1/2 ), √ (e1/2 ⊗ e1/2 − e−1/2 ⊗ e−1/2 ),
2 2
1 1
√ (e1/2 ⊗ e−1/2 + e−1/2 ⊗ e1/2 ), √ (e1/2 ⊗ e−1/2 − e−1/2 ⊗ e1/2 )
2 2
form a basis in C4 ? Prove or disprove.

Solution 5. We obtain the Bell basis

1 0 1 0
       
1 0 1 1 1  0  1  −1 
v1 = √   , v2 = √   , v3 = √   , v4 = √ 
2 0 2 1 0 1

2 2
1 0 −1 0

which forms an orthonormal basis in C4

Problem 6. Let A be an arbitrary n × n matrix over C. Show that

exp(A ⊗ In ) ≡ exp(A) ⊗ In . (1)

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 36

36 Problems and Solutions

Solution 6. Using the expansion

∞
X (A ⊗ In )k
exp(A ⊗ In ) =
k!
k=0
1 1 1
= In ⊗ In + (A ⊗ In ) + (A ⊗ In )2 + (A ⊗ In )3 + · · ·
1! 2! 3!
and (A ⊗ In )k = Ak ⊗ In , k ∈ N we find identity (1).

Problem 7. Let A, B be arbitrary n × n matrices over C. Let In be the

n × n unit matrix. Show that

exp(A ⊗ In + In ⊗ B) ≡ exp(A) ⊗ exp(B).

Solution 7. The proof of this identity relies on [A ⊗ In , In ⊗ B] = 0n2

and

(A ⊗ In )r (In ⊗ B)s ≡ (Ar ⊗ In )(In ⊗ B s ) ≡ Ar ⊗ B s , r, s ∈ N.

Thus
∞
X (A ⊗ In + In ⊗ B)j
exp(A ⊗ In + In ⊗ B) =
j=0
j!
∞ Xj  
X 1 j
= (A ⊗ In )k (In ⊗ B)j−k
j=0
j! k
k=0
j
∞ X  
X 1 j
= (Ak ⊗ B j−k )
j=0 k=0
j! k
 
∞ ∞
!
j
X A X Bk
=   ⊗
j=0
j! k!
k=0

= exp(A) ⊗ exp(B).

Problem 8. Let A and B be arbitrary n × n matrices over C. Prove or

disprove the equation
eA⊗B = eA ⊗ eB .

Solution 8. Obviously this is not true in general. For example, let

A = B = In . Then
eA⊗B = eIn2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 37

Kronecker Product and Tensor Product 37

and
eA ⊗ eB = eIn ⊗ eIn 6= eIn2 .

Problem 9. Let A be an m × m normal matrix and B be an n × n normal

matrix. The underlying field is C. The eigenvalues and eigenvectors of A
are given by λ1 , λ2 , . . . , λm and u1 , u2 , . . . , um . The eigenvalues and
eigenvectors of B are given by µ1 , µ2 , . . . , µn and v1 , v2 , . . . , vn . Let 1 ,
2 and 3 be real parameters. Find the eigenvalues and eigenvectors of the
matrix
1 A ⊗ B + 2 A ⊗ In + 3 Im ⊗ B.

Solution 9. Let x ∈ Cm and y ∈ Cn . Then we have

(A ⊗ B)(x ⊗ y) = (Ax) ⊗ (By),

(A ⊗ In )(x ⊗ y) = (Ax) ⊗ y, (Im ⊗ B)(x ⊗ y) = x ⊗ (By).

Thus the eigenvectors of the matrix are

uj ⊗ vk , j = 1, 2, . . . , m k = 1, 2, . . . , n.

The corresponding eigenvalues are given by 1 λj µk + 2 λj + 3 µk .

Problem 10. Let A, B be n × n matrices over C. The n × n matrices

form a vector space. A scalar product can be defined as

hA, Bi := tr(AB ∗ ).

This provides a Hilbert space. The scalar product implies a norm

kAk2 = hA, Ai = tr(AA∗ ).

This norm is called the Hilbert-Schmidt norm.

(i) Consider the Dirac matrices
1 0 0 0 0 0 0 1
   
0 1 0 0   0 0 1 0
γ0 :=  , γ1 :=  .
0 0 −1 0 0 −1 0 0
0 0 0 −1 −1 0 0 0
Calculate the scalar product hγ0 , γ1 i. Discuss.
(ii) Let U be a unitary n × n matrix. Find the scalar product hU A, U Bi.
(iii) Let C, D be m × m matrices over C. Find hA ⊗ C, B ⊗ Di.

Solution 10. (i) We find hγ0 , γ1 i = tr(γ0 γ1∗ ) = 0. Thus γ0 and γ1 are
orthogonal to each other.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 38

38 Problems and Solutions

(ii) Since

tr(U A(U B)∗ ) = tr(U AB ∗ U ∗ ) = tr(U ∗ U AB ∗ ) = tr(AB ∗ )

where we used the cyclic invariance for matrices, we find that

hU A, U Bi = hA, Bi.

Thus the scalar product is invariant under unitary transformations.

(iii) Since

tr((A ⊗ C)(B ⊗ D)∗ ) = tr((A ⊗ C)(B ∗ ⊗ D∗ )) = tr((AB ∗ ) ⊗ (CD∗ ))

= tr(AB ∗ )tr(CD∗ )

we find hA ⊗ C, B ⊗ Di = hA, BihC, Di.

Problem 11. Let T be the 4 × 4 matrix

 
X3
T := I2 ⊗ I2 + t j σj ⊗ σj 
j=1

where σj , j = 1, 2, 3 are the Pauli spin matrices and −1 ≤ tj ≤ +1,

j = 1, 2, 3. Find the matrix T 2 .

Solution 11. We have

3
X 3 X
X 3
T 2 = I2 ⊗ I2 + 2 t j σj ⊗ σj + tj tk σj σk ⊗ σj σk .
j=1 j=1 k=1

Since
σ1 σ2 = iσ3 , σ2 σ1 = −iσ3 , σ2 σ3 = iσ1 ,
σ3 σ2 = −iσ1 , σ3 σ1 = iσ2 , σ1 σ3 = −iσ2
and σ12 = I2 , σ22 = I2 , σ32 = I2 , we find
3
X 3
X
tj tk σj σk ⊗σj σk ≡ I2 ⊗I2 t2j −2(t1 t2 σ3 ⊗σ3 +t2 t3 σ1 ⊗σ1 +t3 t1 σ2 ⊗σ2 ).
j,k=1 j=1

Therefore
 
3
X
T 2 = (I2 ⊗ I2 ) 1 + t2j 
j=1

+2(t1 − t2 t3 )σ1 ⊗ σ1 + 2(t2 − t3 t1 )σ2 ⊗ σ2 + 2(t3 − t1 t2 )σ3 ⊗ σ3 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 39

Kronecker Product and Tensor Product 39

Problem 12. Let { |0i, |1i, . . . , |n − 1i } be an orthonormal basis in the

Hilbert space Cn . Is
 
n−2
1 X
|ψi = √ |ji ⊗ |j + 1i + |n − 1i ⊗ |0i
n j=0

independent of the chosen orthonormal basis? Prove or disprove.

Solution 12. Consider the special case R2 . Let

   
1 0
|0i = , |1i = .
0 1

Thus
0
 
       
1 1 0 1 0 1 1 1
|ψi = √ ⊗ +√ ⊗ = √  .
2 0 1 2 1 0 2 1
0

Now let    
1 1 1 1
|0i = √ , |1i = √ .
2 1 2 −1
Then
1
 
       
1 1 1 1 1 1 1 1 1  0 
|ψi = √ ⊗ +√ ⊗ =√  .
2 2 1 −1 2 2 −1 1 2 0
−1

Thus, |ψi depends on the chosen basis.

Problem 13. In the product Hilbert space C4 ∼

= C2 ⊗ C2 the Bell states
are given by
1 1
|Φ+ i = √ (|0i ⊗ |0i + |1i ⊗ |1i), |Φ− i = √ (|0i ⊗ |0i − |1i ⊗ |1i),
2 2
1 1
|Ψ+ i = √ (|0i ⊗ |1i + |1i ⊗ |0i), |Ψ− i = √ (|0i ⊗ |1i − |1i ⊗ |0i)
2 2
and form an orthonormal basis in C4 . Here, { |0i, |1i } is an arbitrary
orthonormal basis in the Hilbert space C2 . Let
 iφ   iφ 
e cos(θ) −e sin(θ)
|0i = , |1i = .
sin(θ) cos(θ)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 40

40 Problems and Solutions

(i) Find |Φ+ i, |Φ− i, |Ψ+ i, |Ψ− i for this basis.

(ii) Consider the special case when φ = 0 and θ = 0.

Solution 13. (i) We obtain

 2iφ   2iφ
e e cos(2θ)

1  0  iφ
1  e sin(2θ) 
|Φ+ i = √  , |Φ− i = √  iφ ,
2 0 2 e sin(2θ)
1 − cos(2θ)
 2iφ
−e sin(2θ) 0
  
1  eiφ cos(2θ)  1  eiφ 
|Ψ+ i = √  iφ , |Ψ− i = √  iφ  .
2 e cos(2θ) 2 −e
sin(2θ) 0
(ii) If we choose φ = 0 and θ = 0 which simply means we choose the
standard basis for |0i and |1i (i.e. |0i = (1 0)T and |1i = (0 1)T ), we find
that the Bell states take the form
1 1
   
1 0 1  0 
|Φ+ i = √   , |Φ− i = √  ,
2 0 2 0
1 −1
0 0
   
1 1 1  1 
|Ψ+ i = √   , |Ψ− i = √  .
2 1 2 −1
0 0

Problem 14. Let HA and HB be two p-dimensional Hilbert spaces over

C, where p is a prime number. Let

{ |0iA , |1iA , . . . , |(p − 1)iA }, { |0iB , |1iB , . . . , |(p − 1)iB }

be orthonormal bases in these Hilbert spaces. We define the states

p−1
1 X
a b
|ψ(a, b)i := (Ip ⊗ X Z ) √ |jiA ⊗ |jiB
p j=0

in the Hilbert space HA ⊗ HB , where a, b ∈ { 0, 1, . . . , p − 1 }. The p × p

matrices X and Z are defined as

X|ji = |j + 1 mod pi, Z|ji = ω j |ji, j = 0, 1, . . . , p

with a complex primitive pth root ω of 1 and { |0i, |1i, . . . , |p − 1i } is the

orthonormal basis given above for the Hilbert space HB . Calculate the
states |ψ(0, 0)i and |ψ(1, 1)i.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 41

Kronecker Product and Tensor Product 41

Solution 14. Since

X 0 = Z 0 = Ip , Ip |jiA = |jiA , Ip |jiB = |jiB

we obtain
p−1
1 X
|ψ(0, 0)i = √ |jiA ⊗ |jiB .
p j=0

Using

(Ip ⊗ XZ)(|jiA ⊗ |jiB = |jiA ⊗ (XZ|jiB ) = |jiA ⊗ ω j X|jiB

= |jiA ⊗ ω j |jB + 1 mod pi
= ω j |jiA ⊗ |jB + 1 mod pi

we find
p−1
1 X j
|ψ(1, 1)i = √ ω |jiA ⊗ |jB + 1 mod pi.
p j=0

The states |ψ(a, b)i are maximally entangled states in the Hilbert space
HA ⊗ HB .

Problem 15. Consider the Pauli matrices σ1 and σ2 and the GHZ state
           
1 1 1 1 0 0 0
|ψi = √ ⊗ ⊗ + ⊗ ⊗ .
2 0 0 0 1 1 1

(i) Show that |ψi is an eigenvector of the operator σ2 ⊗ σ2 ⊗ σ1 . What is

the eigenvalue?
(ii) Is
           
1 1 1 1 0 0 0
|φi = √ ⊗ ⊗ + eiφ ⊗ ⊗
2 0 0 0 1 1 1

an eigenvector of σ2 ⊗ σ2 ⊗ σ1 ?

Solution 15. (i) We have

               
1 0 0 1 1 0 0 −i
σ1 = , σ1 = , σ2 = , σ2 = .
0 1 1 0 0 i 1 0

Thus
      
1 1 1 1
(σ2 ⊗ σ2 ⊗ σ1 )|ψi = √ σ2 ⊗ σ2 ⊗ σ1
2 0 0 0
     
0 0 0
+ σ2 ⊗ σ2 ⊗ σ1
1 1 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 42

42 Problems and Solutions

           
1 0 0 0 −i −i 1
=√ ⊗ ⊗ + ⊗ ⊗
2 i i 1 0 0 0
           
1 0 0 0 1 1 1
= −√ ⊗ ⊗ + ⊗ ⊗
2 1 1 1 0 0 0
= −|ψi.

Consequently, the eigenvalue is −1.

(ii) We find using the calculation from (i)
            
1 1 1 1 0 0 0
(σ2 ⊗σ2 ⊗σ1 )|φi = − √ eiφ ⊗ ⊗ + ⊗ ⊗ .
2 0 0 0 1 1 1

Thus, in general |φi is not an eigenvector of σ2 ⊗ σ2 ⊗ σ1 , for example if

φ = π/4.

Problem 16. Consider the three-qubit GHZ state

1
|ψi = √ (|0i ⊗ |0i ⊗ |0i + |1i ⊗ |1i ⊗ |1i)
2
with the standard basis
   
1 0
|0i = , |1i = .
0 1

Let σ1 , σ2 be the Pauli spin matrices.

(i) Calculate the expectation values

hψ|(σ2 ⊗ σ2 ⊗ σ2 )|ψi, hψ|(σ1 ⊗ σ1 ⊗ σ2 )|ψi,

hψ|(σ1 ⊗ σ2 ⊗ σ1 )|ψi, hψ|(σ2 ⊗ σ1 ⊗ σ1 )|ψi.

(ii) Calculate the expectation values

hψ|(σ1 ⊗ σ2 ⊗ σ2 )|ψi, hψ|(σ2 ⊗ σ1 ⊗ σ2 )|ψi,

hψ|(σ2 ⊗ σ2 ⊗ σ1 )|ψi, hψ|(σ1 ⊗ σ1 ⊗ σ1 )|ψi.

Solution 16. (i) We find

hψ|(σ2 ⊗ σ2 ⊗ σ2 )|ψi = 0, hψ|(σ1 ⊗ σ1 ⊗ σ2 )|ψi = 0,

hψ|(σ1 ⊗ σ2 ⊗ σ1 )|ψi = 0, hψ|(σ2 ⊗ σ1 ⊗ σ1 )|ψi = 0.

(ii) We obtain

hψ|(σ1 ⊗ σ2 ⊗ σ2 )|ψi = −1, hψ|(σ2 ⊗ σ1 ⊗ σ2 )|ψi = −1,

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 43

Kronecker Product and Tensor Product 43

hψ|(σ2 ⊗ σ2 ⊗ σ1 )|ψi = −1, hψ|(σ1 ⊗ σ1 ⊗ σ1 )|ψi = 1.

Problem 17. Consider the Bell state

       
1 1 0 0 1
|Ψ− i = √ ⊗ − ⊗ .
2 0 1 1 0

Let n, m be unit vectors in R3 . Calculate the expectation value

E(n, m) = hΨ− |(n · σ) ⊗ (m · σ)|Ψ− i.

Solution 17. Straightforward calculation yields

E(n, m) = −n1 m1 −n2 m2 −n3 m3 = −n·m = −|n|·|m| cos(θ) = − cos(θm,n )

since |m| = |n| = 1. We write cos(θm,n ) instead of cos(θ) in order to

indicate that θn,m is the angle between the quantization directions m and
n.

Problem 18. Let A1 , A2 , . . . , Am and B1 , B2 , . . . , Bn be two sets

of 4 × 4 matrices over C. Assume that Aj Bk = Bk Aj for all j, k with
j = 1, 2, . . . , m and k = 1, 2, . . . , n. Find two such sets of matrices using
the Kronecker product of 2 × 2 matrices.

Solution 18. If Aj = Cj ⊗ I2 , Bk = I2 ⊗ Dk , where Cj , Dk are arbitrary

2 × 2 matrices we have

[Aj , Bk ] = [Cj ⊗ I2 , I2 ⊗ Dk ] = Cj ⊗ Dk − Cj ⊗ Dk = 04

for all j, k.

Problem 19. Let A, B be n × n hermitian matrices over C and

H = ~ω(A ⊗ B)

be a Hamilton operator, where ~ is the Planck constant and ω the frequency.

The Heisenberg equation of motion for the operator B ⊗ A is given by
d(B ⊗ A)
i~ = [B ⊗ A, H](t).
dt
(i) Assume that [A, B] = 0n . Simplify the Heisenberg equation of motion
using this condition.
(ii) Assume that [A, B]+ = 0n Simplify the Heisenberg equation of motion
using this condition. Give an example for such matrices.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 44

44 Problems and Solutions

(iii) Assume that [A, B]+ = In . Simplify the Heisenberg equation of motion
using this condition. Give an example for such matrices.

Solution 19. (i) We have

[B ⊗ A, Ĥ](t) = ~ω[B ⊗ A, A ⊗ B](t)

= ~ω((B ⊗ A)(A ⊗ B) − (A ⊗ B)(B ⊗ A))(t)
= ~ω((BA) ⊗ (AB) − (AB) ⊗ (BA))(t).

In general, (AB) ⊗ (BA) 6= (BA) ⊗ (AB). If [A, B] = 0n we find

d(B ⊗ A)
i~ = 0n2 .
dt
Thus B ⊗ A is a constant of motion.
(ii) From [A, B]+ = 0n we have AB = −BA. Thus we also have

d(B ⊗ A)
i~ = 0n2
dt
in this case. An example are the Pauli spin matrices. For example, let
A = σ1 and B = σ2 . Then [A, B]+ = 02 . Another example is given by
Fermi operators. They have the matrix representation (j = 1, 2, . . . , N )
 
1
c†j = σ3 ⊗ · · · ⊗ σ3 ⊗ σ+ ⊗ I2 ⊗ · · · ⊗ I2
2
 
1
cj = σ3 ⊗ · · · ⊗ σ3 ⊗ σ− ⊗ I2 · · · ⊗ I2
2
where σ+ and σ− appears in the j-th place and where σ+ := σ1 + iσ2 ,
σ− := σ1 − iσ2 . We have [c†k , cj ]+ = δjk I and [c†k , c†j ]+ = [ck , cj ]+ = 0.
(iii) Since BA = −AB + In we find

d(B ⊗ A)
i~ = ~ω(In ⊗ (AB) − (AB) ⊗ In )(t).
dt

Problem 20. The four Bell states with spin J (J = 1/2, 1, 3/2, 2, . . .) are
given by
2J
1 X
|B1 i = √ |ki ⊗ |ki
2J + 1 k=0
2J
1 X
|B2 i = √ (−1)k |ki ⊗ |ki
2J + 1 k=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 45

Kronecker Product and Tensor Product 45

2J
1 X
|B3 i = √ |ki ⊗ |2J − ki
2J + 1 k=0
2J
1 X
|B4 i = √ (−1)k |ki ⊗ |2J − ki.
2J + 1 k=0

For J = 1/2 we obtain the standard Bell basis in C4 . Note that for J = 1/2
the four Bell states form an orthonormal basis in C4 . Show that for spin
J = 1 the Bell states are linearly dependent.

Solution 20. For J = 1 the Bell states are

1
|B1 i = √ (|0i ⊗ |0i + |1i ⊗ |1i + |2i ⊗ |2i)
3
1
|B2 i = √ (|0i ⊗ |0i − |1i ⊗ 1i + |2i ⊗ |2i)
3
1
|B3 i = √ (|0i ⊗ |2i + |1i ⊗ |1i + |2i ⊗ |0i)
3
1
|B4 i = √ (|0i ⊗ |2i − |1i ⊗ |1i + |2i ⊗ |0i).
3
Let c1 , c2 , c3 , c4 ∈ C. Then the equation
c1 |B1 i + c2 |B2 i + c3 |B3 i + c4 |B4 i = 0
provides the solution c2 = −c1 , c3 = −c1 , c4 = c1 with c1 arbitrary. Thus
the four states are not linearly independent. We also have
1 1 1
hB1 |B2 i = , hB1 |B3 i = , hB1 |B4 i = − ,
3 3 3
1 1 1
hB2 |B3 i = − , hB2 |B4 i = , hB3 |B4 i = .
3 3 3

Problem 21. Consider the state in C4

1
X
|ψi = cj1 j2 |j1 i ⊗ |j2 i
j1 ,j2 =0

where    
1 0
|0i = , |1i = .
0 1
Let S, T be 2 × 2 matrices over C
   
s11 s12 t11 t12
S= , T =
s21 s22 t21 t22
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 46

46 Problems and Solutions

where det(S) = 1, det(T ) = 1. This means that S and T are elements of

the Lie group SL(2, C). Let
1
X 1
X
dj1 j2 |j1 i ⊗ |j2 i = (S ⊗ T ) cj1 j2 |j1 i ⊗ |j2 i.
j1 ,j2 =0 j1 ,j2 =0

Show that
d00 d11 − d01 d10 = c00 c11 − c01 c10 . (1)
Owing to (1) the quantity c00 c11 − c01 c10 is called an invariant.

Solution 21. Since

s11 t11 s11 t12 s12 t11 s12 t12
 
 s11 t21 s11 t22 s12 t21 s12 t22 
S⊗T =
s21 t11 s21 t12 s22 t11 s22 t12

s21 t21 s21 t22 s22 t21 s22 t22

we have
s11 t11 s11 t12 s12 t11 s12 t12
       
s t  s t  s t  s t 
(S⊗T )|ψi = c00  11 21 +c01  11 22 +c10  12 21 +c11  12 22  .
s21 t11 s21 t12 s22 t11 s22 t12
s21 t21 s21 t22 s22 t21 s22 t22

Thus

d00 = c00 s11 t11 + c01 s11 t12 + c10 s12 t11 + c11 s12 t12
d01 = c00 s11 t21 + c01 s11 t22 + c10 s12 t21 + c11 s12 t22
d10 = c00 s21 t11 + c01 s21 t12 + c10 s22 t11 + c11 s22 t12
d11 = c00 s21 t21 + c01 s21 t22 + c10 s22 t21 + c11 s22 t22 .

From det(S) = 1 and det(T ) = 1 it follows that s11 s22 − s12 s21 = 1 and
t11 t22 − t12 t21 = 1. Thus

s11 s22 t11 t22 + s12 s21 t12 t21 − s11 s22 t12 t21 − s12 s21 t11 t22 = 1.

Using this result we obtain (1).

Problem 22. Let I2 be the 2 × 2 identity matrix and let σ1 , σ2 and σ3

be the Pauli spin matrices.
(i) Let A be an m × m matrix over C and B be an n × n matrix over C.
Find all solutions for m, n, A and B satisfying

1 1 1 1
I2 ⊗ I2 + σ1 ⊗ σ1 + σ2 ⊗ σ2 + σ3 ⊗ σ3 = A ⊗ B.
2 2 2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 47

Kronecker Product and Tensor Product 47

(ii) Let C, D, E, and F be 2 × 2 matrices over C. Find all solutions for C,

D, E and F satisfying
1 1 1 1
I2 ⊗ I2 + σ1 ⊗ σ1 + σ2 ⊗ σ2 + σ3 ⊗ σ3 = C ⊗ D + E ⊗ F.
2 2 2 2
(iii) Find the set stabilized by S = { σ1 ⊗ I2 , I2 ⊗ σ1 } i.e. find

u ∈ C4 | ∀ A ∈ S : Au = u .


Solution 22. (i) Straightforward calculation yields

1 0 0 0
 
1 1 1 1 0 0 1 0
I2 ⊗ I2 + σ1 ⊗ σ1 + σ2 ⊗ σ2 + σ3 ⊗ σ3 =   = A ⊗ B.
2 2 2 2 0 1 0 0
0 0 0 1
Since A ⊗ B is an mn × mn matrix we have mn = 4. For m = n = 2 we
consider    
a1 a2 b1 b2
A= , B= .
a3 a4 b3 b4
Thus we have a1 b1 = 1, a1 b4 = 0 and a4 b4 = 1. Consequently a1 6= 0 and
b4 6= 0, so that a1 b4 6= 0. Thus m = n = 2 does not yield a solution.

For m = 1 and n = 4 we find the solution

1 0 0 0
 
1 0 0 1 0
A = (a), B=  , a ∈ C/{0}.
a 0 1 0 0
0 0 0 1
For m = 4 and n = 1 we find the solution
1 0 0 0
 
1 0 0 1 0
A=  , B = (b), b ∈ C/{0}.
b 0 1 0 0
0 0 0 1
(ii) Clearly C and E, and D and F must be linearly independent, otherwise
we would have the case m = n = 2 discussed in (a) which has no solution.
Let    
c1 c2 e1 e2
C= , E= .
c3 c4 e3 e4
We have to satisfy the equations
   
1 0 0 0
c1 D + e1 F = , c2 D + e2 F = ,
0 0 1 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 48

48 Problems and Solutions

   
0 1 0 0
c3 D + e3 F = , c4 D + e4 F = .
0 0 0 1
It follows that {D, F } should be a basis for the 2 × 2 matrices over C,
however the span of {D, F } is a two dimensional space whereas the 2 × 2
matrices over C forms a four dimensional space. Thus we have a contradic-
tion. There are no solutions.
(iii) Thus the vector u must simultaneously be an eigenvector of σ1 ⊗I2 and
I2 ⊗ σ1 with eigenvalue 1. The eigenspace corresponding to the eigenvalue
1 from σ1 ⊗ I2 yields
       
1 1 1 0
u=α ⊗ +β ⊗
1 0 1 1

where α, β ∈ C. We must now satisfy

       
1 1 1 0
(I2 ⊗ σ1 )u = α(I2 ⊗ σ1 ) ⊗ + β(I2 ⊗ σ1 ) ⊗
1 0 1 1
       
1 0 1 1
=α ⊗ +β ⊗
1 1 1 0
=u

so that α = β. Thus
       
1 1 1 0
u=α ⊗ +α ⊗ , α ∈ C.
1 0 1 1

N
Problem 23. Let N ≥ 1. Consider the Hilbert space C2 . The (N + 1)
Dicke states are defined by
1
|N/2, ` − N/2i := p (|0i ⊗ · · · ⊗ |0i ⊗ |1i ⊗ · · · ⊗ |1i +permutations)
NC
`
| {z } | {z }
` N −`

where ` = 0, 1, . . . , N and
N
C` = N !/(`!(N − `)!).

Write down the Dicke states for N = 2 and N = 3. Which of the states are
entangled?

Solution 23. For N = 2 we have the three states in the Hilbert space C4
1
|1, −1i = |1i ⊗ |1i, |1, 0i = √ (|0i ⊗ |1i + |1i ⊗ |0i), |1, 1i = |0i ⊗ |0i.
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 49

Kronecker Product and Tensor Product 49

The first and the last states are product states. The second state is a Bell
state and fully entangled. For N = 3 we find the four states in the Hilbert
space C8

3 3
 , − i = |1i ⊗ |1i ⊗ |1i
2 2

3 1
 , − i = √1 (|0i ⊗ |1i ⊗ |1i + |1i ⊗ |0i ⊗ |1i + |1i ⊗ |1i ⊗ |0i)
2 2 3

3 1 1
 , i = √ (|0i ⊗ |0i ⊗ |1i + |0i ⊗ |1i ⊗ |0i + |1i ⊗ |0i ⊗ |0i)
2 2 3

3 3
 , i = |0i ⊗ |0i ⊗ |0i.
2 2

Obviously the first and last states are product states. The other two states
are entangled.

Problem 24. Can we find 2 × 2 matrices A, B, C with det(A) = 1,

det(B) = 1 and det(C) = 1 such that

0 1
   
1 0
1 0
   
1 0
  1 
0

√   = (A ⊗ B ⊗ C) √   ?
3 1 2 0
0 0
   
0 0
   
0 1

On the left-hand side we have the W state and on the right-hand side we
have the GHZ state.

Solution 24. Let

 
a11 a12
A= , det(A) = a11 a22 − a12 a21 = 1
a21 a22

etc. Thus the condition yields 8 equations and we have the three constraints

a11 a22 − a12 a21 = 1, b11 b22 − b12 b21 = 1, c11 c22 − c12 c21 = 1.

There is no solution for this system.

Problem 25. Consider the vector

1 T
v= (1 0 1 1 0 1 ) ∈ C6 .
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 50

50 Problems and Solutions

Find a Schmidt decomposition of v over C6 = C2 ⊗ C3 and over C6 =

C3 ⊗ C2 .

Solution 25. Over C2 ⊗ C3 we have

     
  1   1   1
1 1 1 0 1 1 1
v= ⊗ 0 + ⊗ 0 = √ ⊗ √ 0
2 0 2 1 2 1 2 1
1 1

which is trivially a Schmidt decomposition of v. Thus, over C2 ⊗ C3 , v has

Schmidt rank 1. Over C3 ⊗ C2 we have
     
1   0   0  
1  1 1  1 1  0
v= 0 ⊗ + 1 ⊗ + 0 ⊗ .
2 0 2 1 2 1
0 0 1

Identifying C3 with the columns of a matrix and C2 with the rows of a

matrix, we rearrange v to yield the matrix
 
1 0
1
A= 1 1
2
0 1

which has the singular value decomposition

 1  √
√ √1 √1 3
 !∗
6 2 3
2 0 √1 √1
 √2 0 − √13   0 1  2 2
A= 6 .

2 √1 − √12
1 1 1 0 0 2
√ −√6
√
2 3

From the singular value decomposition we obtain the Schmidt decomposi-

tion
√
   
1   1  
3 1   1 1 1 1 1 1
v= ·√ 2 ⊗√ + · √  0 ⊗ √ .
2 6 1 2 1 2 2 −1 2 −1

Thus, over C3 ⊗ C2 , v has Schmidt rank 2.

Problem 26. Let A be an m × m hermitian matrix and let B be an

n × n hermitian matrix. Then A ⊗ B, A ⊗ In , Im ⊗ B are also hermitian
matrices, where Im is the m × m identity matrix. Let 1 , 2 and 3 be real
parameters. Consider the Hamilton operator

H = ~ω(1 A ⊗ B + 2 A ⊗ In + 3 Im ⊗ B).

The partition function Z(β) is given by Z(β) = tr(exp(−βH)), where H

is the (hermitian) Hamilton operator and tr denotes the trace. From the
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 51

Kronecker Product and Tensor Product 51

partition function we obtain the Helmholtz free energy, entropy and specific
heat.
(i) Calculate Z(β) for the Hamilton operator given above.
(ii) Consider the special case that n = m = 2 and A, B are any of the Pauli
spin matrices σ1 , σ2 , σ3 .

Solution 26. Since A is an m × m matrix we can find a m × m unitary

matrix UA such that A e = U ∗ AUA is a diagonal matrix. We set diag(A) e =
A
(λ1 , λ2 , . . . , λm ). Analogously for the n × n hermitian matrix B we find a
n×n hermitian matrix UB such that B e = U ∗ BUB is a diagonal matrix. We
B
set diag(B) e = (µ1 , µ2 , . . . , µn ). Since A and B are hermitian the diagonal
elements of A e and B
e are real. Since UA and UB are unitary matrices we
find that UA ⊗ UB is also a unitary matrix and (UA ⊗ UB )∗ = UA∗ ⊗ UB∗ .
Now we find
tr(e−βH ) = tr((UA∗ ⊗ UB∗ )e−βH (UA ⊗ UB ))
∗ ∗
= tre−β(UA ⊗UB )H(UA ⊗UB )
∗ ∗ ∗ ∗
= tre−β~ω(1 (UA AUA )⊗(UB BUB )+2 (UA AUA )⊗In +3 (Im ⊗(UB BUB ))
= tre−β~ω(1 A⊗B+2 A⊗In +3 Im ⊗B)
e e e e

Xm Xn
= e−β~ω(1 λj µk +2 λj +3 µk ) .
j=1 k=1

This calculation can be extended straightforward to the matrix

A1 ⊗ A2 ⊗ A3 + A1 ⊗ In2 ⊗ In3 + In1 ⊗ A2 ⊗ In3 + In1 ⊗ In2 ⊗ A3
and so on, where A1 , A2 , A3 are n1 × n1 , n2 × n2 , n3 × n3 matrices,
respectively.
(ii) The eigenvalues of any Pauli spin matrix are +1 and −1. Thus for any
combination σj ⊗ σk for A ⊗ B we find
Z(β) = e−β~ω1 2 cosh(β~ω(2 + 3 )) + eβ~ω1 2 cosh(β~ω(2 − 3 )).

Problem 27. (i) Let { |0i, |1i, . . . , |n − 1i } be an orthonormal basis in

Cn . Is
n−1
1 X
|ψi := √ |ji ⊗ |ji
n j=0
independent of the chosen orthonormal basis?
(ii) Find the density matrix |ψihψ|.
(iii) Consider the linear operator
n−1 n−1
1 XX
Π := (|jki − |kji)(hjk| − hkj|)
4 j=0 k=0
k6=j
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 52

52 Problems and Solutions

where we used the short-cut notation |jki ≡ |ji ⊗ |ki. Calculate Π∗ and
Π2 . What is the use of this operator?

Solution 27. (i) Let { |φ0 i, |φ1 i, . . . , |φn−1 i } be an orthonormal basis in

Cn . Then we have the expansion for the state |ji
n−1
X
|ji = hj|φk i|φk i.
k=0

Thus |ψi can be written as

n−1 n−1
! n−1
!!
1 X X X
|ψi = √ hj|φk i|φk i ⊗ hj|φl i|φl i
n j=0
k=0 l=0
n−1 n−1 n−1
1 XXX
=√ hφk |jihj|φl i|φk i ⊗ |φl i
n j=0
k=0 l=0
 
n−1 n−1 n−1
1 X X X
=√ hφk |jihj|φl i |φk i ⊗ |φl i
n j=0
k=0 l=0

where we used hj|φk i = hφk |ji. Note that for the sum
n−1
X
hφk |jihj|φl i
j=0

we cannot apply Parseval’s relation. Parseval’s relation would apply to

n−1
X
hφk |jihφl |ji = hφk |φl i = δkl .
j=0

Thus the Bell state |ψi is dependent on the chosen basis. However, if
all scalar products hj|φk i are real numbers then |ψi is independent of the
chosen basis.
(ii) We have
n−1 n−1 n−1 n−1
1 XX 1 XX
|ψihψ| = (|ji ⊗ |ji) (hk| ⊗ hk|) = |jihk| ⊗ |jihk|.
n j=0 n j=0
k=0 k=0

(iii) Clearly, Π∗ = Π. Furthermore

n−1 n−1
1 XXX X
Π2 = (|jki − |kji)(hjk| − hkj|)(|lmi − |mli)(hlm| − hml|)
16 j=0
j6=k l=0 l6=m
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 53

Kronecker Product and Tensor Product 53

n−1 n−1
1 XXX X
= (|jki − |kji)(2δjl δkm − 2δjm δlk )(hlm| − hml|)
16 j=0
j6=k l=0 l6=m
= Π.

Thus Π is a projection matrix. It projects onto the space spanned by

 
1
√ (|jki − |kji) : j, k ∈ {0, 1, . . . , n − 1}, k > j .
2

Problem 28. Let HA and HB be two finite-dimensional Hilbert spaces.

The Schmidt rank of a linear operator L : HA ⊗ HB → HA ⊗ HB over
HA ⊗ HB is the smallest non-negative integer Sch(L, HA , HB ) such that L
can be written as
Sch(L,HA ,HB )
X
L= Lj,A ⊗ Lj,B
j=1

where Lj,A : HA → HA and Lj,B : HB → HB are linear operators.

Let {|0i, |1i} denote an orthonormal basis in C2 . Find the Schmidt rank
Sch(UCN OT , C2 , C2 ) and Sch(USW AP , C2 , C2 ) where

UCN OT = |00ih00| + |01ih01| + |11ih10| + |10ih11|

USW AP = |00ih00| + |10ih01| + |01ih10| + |11ih11|.

Solution 28. We note that

UCN OT = |0ih0| ⊗ I2 + |1ih1| ⊗ UN OT

where UN OT := |0ih1| + |1ih0|. In other words

0 < Sch(UCN OT , C2 , C2 ) ≤ 2.

Now suppose UCN OT can be written as the product A ⊗ B where

A := a0 |0ih0| + a1 |0ih1| + a2 |1ih0| + a3 |1ih1|

B := b0 |0ih0| + b1 |0ih1| + b2 |1ih0| + b3 |1ih1|.

This yields the conditions a0 b0 = 1, a0 b1 = 0 and a3 b1 = 1. These equations

are inconsistent, i.e.
Sch(UCN OT , C2 , C2 ) 6= 1.
Thus
Sch(UCN OT , C2 , C2 ) = 2.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 54

54 Problems and Solutions

The operator USW AP has the eigenvalue 1 (three times) with corresponding
orthonormal eigenvectors
 
1
|00i, |11i, √ (|01i + |10i)
2
and the eigenvalue −1 with corresponding eigenvector √1 (|01i − |10i).
2
Defining
1 1
|φ1 i := √ (|01i + |10i), |φ2 i := √ (|01i − |10i)
2 2
we find that

USW AP := |00ih00| + |φ1 ihφ1 | − |φ2 ihφ2 | + |11ih11|

where {|00i, |φ1 i, |φ2 i, |11i} forms an orthonormal basis in C4 . In this basis
USW AP is the diagonal matrix
1 0 0 0
 
0 1 0 0
USW AP =  .
0 0 −1 0
0 0 0 1
Clearly, the matrices

|00ih00|, |11ih11|, |φ1 ihφ1 | and |φ2 ihφ2 |

are linearly independent. Thus Sch(USW AP , C2 , C2 ) = 4.

Problem 29. The operator-Schmidt decomposition of a linear operator Q

acting in the product Hilbert space H = H1 ⊗ H2 of two finite-dimensional
Hilbert spaces (dim(H1 ) = m, dim(H2 ) = n) with H1 = Cm and H2 = Cn
can be constructed as follows. Let X, Y be d × d matrices over C. Then
we can define a scalar product or inner product hX, Y i := tr(XY ∗ ). Using
this inner product we can define an orthonormal set of d × d matrices
{ Xj : j = 1, 2, . . . , d2 } which satisfies the condition

hXj , Xk i = tr(Xj Xk∗ ) = δjk .

Thus we can write the matrix Q as

2 2
m X
X n
Q= cjk Aj ⊗ Bk
j=1 k=1

where { Aj : j = 1, 2, . . . , m2 } and { Bk : k = 1, 2, . . . , n2 } are fixed

orthonormal bases of m × m and n × n matrices in the Hilbert spaces Cm
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 55

Kronecker Product and Tensor Product 55

and Cn respectively, and cjk are complex coefficients. Thus C = (cjk ), with
j = 1, 2, . . . , m2 and k = 1, 2, . . . , n2 is an m2 × n2 matrix. The singular
value decomposition theorem states that the matrix C can be written as

C = U ΣV ∗

where U is an m2 × m2 unitary matrix, V is an n2 × n2 unitary matrix and

Σ is an m2 × n2 diagonal matrix. The matrix Σ is of the form
s1 . . . 0
 
 .. . . .. 
 . . . 
 0 . . . sn2 
 
Σ= .
 0 ... 0 
 . . .. 
 .. .. . 
0 ... 0
It is assumed that C, U and V are calculated in orthonormal bases, for
example the standard basis. Thus we obtain
2 2 2
m X
X n X
n
Q= Uj` s` V`k Aj ⊗ Bk
j=1 k=1 `=1

where s` is the `-th diagonal entry of the m2 × n2 diagonal matrix Σ.

Defining
m2
X n2
X
H` := Uj` Aj , K` := V`k Bk
j=1 k=1
2
where ` = 1, 2, . . . , n we find the operator-Schmidt decomposition
2
n
X
Q= s` H` ⊗ K` .
`=1

(i) Consider the CNOT gate

1 0 0 0
 
0 1 0 0
UCN OT = .
0 0 0 1
0 0 1 0
Find the operator-Schmidt decomposition of UCN OT .
(ii) Consider the SWAP operator
1 0 0 0
 
0 0 1 0
USW AP =  .
0 1 0 0
0 0 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 56

56 Problems and Solutions

Find the operator-Schmidt decomposition of USW AP .

(iii) Let
p √  p √ 
Z := 1 − pI2 ⊗ I2 + i pσ1 ⊗ σ1 1 − pI2 ⊗ I2 + i pσ3 ⊗ σ3

where σ1 , σ2 and σ3 are the Pauli spin matrices. Find the operator-Schmidt
decomposition of Z.

Solution 29. (i) We have

       
1 0 1 0 0 0 0 1
UCN OT = ⊗ + ⊗
0 0 0 1 0 1 1 0
   
1 0 0 0
= ⊗ I2 + ⊗ σ1 .
0 0 0 1

(ii) We have
1
USW AP = (I2 ⊗ I2 + σ1 ⊗ σ1 + σ2 ⊗ σ2 + σ3 ⊗ σ3 ).
2
(iii) We have
p h    i
Z = (1−p)I2 ⊗I2 +pσ2 ⊗σ2 + p(1 − p) eiπ/4 σ1 ⊗ σ1 + eiπ/4 σ3 ⊗ σ3 .

Programming Problems

Problem 1. Consider the Hadamard basis in C2

   
1 1 1 1
v1 = √ , v2 = √ .
2 1 2 −1

Apply the Kronecker product to find a basis in C4 .

Solution 1. The following Maxima program

/* hadamardbasis.mac */
v1: matrix([1/sqrt(2)],[1/sqrt(2)]);
v1T: transpose(v1);
v2: matrix([1/sqrt(2)],[-1/sqrt(2)]);
v2T: transpose(v2);
r1: v1T . v1;
r2: v1T . v2;
r3: v2T . v2;
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 57

Kronecker Product and Tensor Product 57

v1v1: kronecker_product(v1,v1);
v1v2: kronecker_product(v1,v2);
v2v1: kronecker_product(v2,v1);
v2v2: kronecker_product(v2,v2);
v1v1T: transpose(v1v1);
v1v2T: transpose(v1v2);
v2v1T: transpose(v2v1);
v2v2T: transpose(v2v2);
r4: v1v1T . v2v2T;
provides the orthonormal basis
1 1 1 1
       
1 1 1  −1  1 1  1  −1 
 , , , .
2 1 2 1 2 −1 2 −1
  
1 −1 −1 1

Problem 2. Let S1 , S2 , S3 be the spin matrices for spin- 12

     
1 0 1 1 0 −i 1 1 0
S1 = , S2 = , S3 = .
2 1 0 2 i 0 2 0 −1
Consider the Hamilton operators

e = Ĥ = S1 ⊗ S1 + S2 ⊗ S2 + S3 ⊗ S3
H
~ω

Ke = Ĥ = S1 ⊗ S2 + S2 ⊗ S3 + S3 ⊗ S1 .
~ω
Find the eigenvalues and eigenvectors of H
e and K.
e

Solution 2. Applying the Maxima program

/* eigenS1S2S3.mac */
I2: matrix([1,0],[0,1]);
S1: matrix([0,1/2],[1/2,0]);
S2: matrix([0,-%i/2],[%i/2,0]);
S3: matrix([1/2,0],[0,-1/2]);
T1: kronecker_product(S1,S1);
T2: kronecker_product(S2,S2);
T3: kronecker_product(S3,S3);
H: T1 + T2 + T3;
EH: eigenvectors(H);
X1: kronecker_product(S1,S2);
X2: kronecker_product(S2,S3);
X3: kronecker_product(S3,S1);
K: X1 + X2 + X3;
EK: eigenvectors(K);
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 58

58 Problems and Solutions

we find for H
e the eigenvalues
3 1
− (1×), (3×)
4 4
with the corresponding eigenvectors
0 1 0 0
       
1  1  0 1 1 0
√  ,  , √  ,
2 −1 0 2 1 0
 
0 0 0 1

and for K
e the eigenvalues
3 1
− (1×), (3×)
4 4
with the corresponding eigenvectors
1 1 0
       
0
1  −1  1 0 1  1  1 
, √  , √  , √ 0 .
2 −i 2 0 0

2 2 1 −1
i i −i

Problem 3. Let M be a 2 × 2 matrix and M T the transpose. Let |0i, |1i

be the standard basis in C2 . Show that
(M ⊗ I2 )(|0i ⊗ |0i + |1i ⊗ |1i) = (I2 ⊗ M T )(|0i ⊗ |0i + |1i ⊗ |1i).

Solution 3. The following Maxima program will do the job

/* M22.mac */
b1: matrix([1],[0]); b2: matrix([0],[1]);
b1b1: kronecker_product(b1,b1); b2b2: kronecker_product(b2,b2);
I2: matrix([1,0],[0,1]);
M: matrix([m11,m12],[m21,m22]);
MT: matrix([m11,m21],[m12,m22]);
T1: kronecker_product(M,I2);
T2: kronecker_product(I2,MT);
R1: T1 . b1b1 + T1 . b2b2;
R2: T2 . b1b1 + T2 . b2b2;
F: R1 - R2;

Do we find the same result if we select the orthonormal basis

   
cos(θ) sin(θ)
, ?
sin(θ) − cos(θ)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 59

Kronecker Product and Tensor Product 59

2.3 Supplementary Problems

Problem 1. Consider the standard basis in the vector space of 2 × 2

matrices
       
1 0 0 1 0 0 0 0
E00 = , E01 = , E10 = , E11 =
0 0 0 0 1 0 0 1

and the mutually unbiased basis

   
1 1 0 1 0 1
µ0 = √ , µ1 = √ ,
2 0 1 2 1 0
   
1 0 −i 1 1 0
µ2 = √ , µ3 = √ .
2 i 0 2 0 −1
Express the Bell matrix

1 0 0 1
 
1 0 1 1 0 
B=√ 
2 0 1 −1 0

1 0 0 −1

with the basis given by µj ⊗ µk (j, k = 0, 1, 2, 3).

Problem 2. The following states form an orthonormal basis in the Hilbert

space C3
     
1 0 1
1 1
|π + i = √  0  , |π 0 i =  1  , |π − i = √  0  .
2 1 0 2 −1

These states play a role for the π-mesons. Show that the states

|π + i ⊗ |π + i, |π − i ⊗ |π − i

1 1
√ (|π + i ⊗ |π 0 i + |π 0 i ⊗ |π + i), √ (|π 0 i ⊗ |π − i + |π − i ⊗ |π 0 i)
2 2
1 1
√ (|π + i ⊗ |π 0 i − |π 0 i ⊗ |π + i), √ (|π + i ⊗ |π − i − |π − i ⊗ |π + i),
2 2
1
√ (|π 0 i ⊗ |π − i − |π − i ⊗ |π 0 i)
2
1
√ (2|π 0 i ⊗ |π 0 i + |π + i ⊗ |π − i + |π − i ⊗ |π + i),
6
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 60

60 Problems and Solutions

1
√ (|π + i ⊗ |π − i + |π − i ⊗ |π + i − |π 0 i ⊗ |π 0 i)
3
form an orthonormal basis in the Hilbert space C9 . Which of these states
are entangled?

Problem 3. Let v1 , v2 , v3 be elements of C2 . Find the conditions on

v1 , v2 , v3 such that

v1 ⊗ v2 ⊗ v3 = v3 ⊗ v2 ⊗ v1 .

Problem 4. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Consider the 8 × 8

matrices
K = σ1 ⊗ σ2 ⊗ σ3 , S = σ1 ⊗ σ1 ⊗ σ1 .
Note that the matrices K and S are unitary and hermitian. Show that
[K, S] = 08 . Show that
1 1
Π1 = (I8 + S), Π2 = (I8 − S)
2 2
are projection matrices and Π1 Π2 = 08 . Show that
1 1
Π3 = (I8 − K), Π4 = (I8 + K)
2 2
are projection matrices.

Problem 5. Let σ1 , σ2 , σ3 be the Pauli spin matrices. We have

σ1 σ2 = iσ3 , σ2 σ3 = iσ1 , σ3 σ1 = iσ2

and
σ2 σ1 = −iσ3 , σ3 σ2 = −iσ1 , σ1 σ3 = −iσ2 .
Show that
[σ1 ⊗ σ1 , σ2 ⊗ σ2 ] = 04
and
[σ1 ⊗ σ1 , σ3 ⊗ σ3 ] = 04 .

Problem 6. Can the Z4 Fourier matrix

1 1 1 1
 
 1 i −1 −i 
1 −1 1 −1
 
1 −i −1 i
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 61

Kronecker Product and Tensor Product 61

be written as the Kronecker product of two 2 × 2 unitary matrices?

Problem 7. Let |ai, |bi be normalized states in Cn and X, Y be n × n

matrices over C. Show that
(ha| ⊗ hb|)((X ⊗ In )(In ⊗ Y ))(|ai ⊗ |bi) = ha|X|aihb|Y |bi.
Note that
(|ai ⊗ |bi)∗ = ha| ⊗ hb|.

Problem 8. (i) Consider the standard basis

   
1 0
|0i = , |1i =
0 1
in C2 . Calculate  
0 1
|0ih0| ⊗ I2 + |1ih1| ⊗ .
1 0
(ii) Consider the orthonormal basis
   
cos(θ) sin(θ)
|0i = , |1i =
sin(θ) − cos(θ)
in C2 . Calculate  
0 1
|0ih0| ⊗ I2 + |1ih1| ⊗ .
1 0
Discuss.

Problem 9. Consider the states in C2

   
cos(θ) sin(θ)
|ψ1 i = , |ψ2 i =
sin(θ) − cos(θ)
which form an orthonormal basis in C2 . Find
(hψ1 | ⊗ hψ2 |)(σj ⊗ I2 + I2 ⊗ σj )(|ψ1 i ⊗ |ψ2 i), j = 1, 2, 3
and discuss the dependence on θ.

Problem 10. Consider the four Bell states

1
|ψ1 i = √ (|0i ⊗ |0i + |1i ⊗ |1i)
2
1
|ψ2 i = √ (|0i ⊗ |1i + |1i ⊗ |0i)
2
1
|ψ3 i = √ (|0i ⊗ |1i − |1i ⊗ |0i)
2
1
|ψ4 i = √ (|0i ⊗ |0i − |1i ⊗ |1i)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 62

62 Problems and Solutions

and the three 4 × 4 matrices

Tj := σj ⊗ I2 + I2 ⊗ σj , j = 1, 2, 3.

Calculate
Tj |ψk i, j = 1, 2, 3 k = 1, 2, 3, 4.
Which of these expressions is an eigenvalue equation? Calculate

hψk |Tj |ψk i.

Problem 11. Let A be an m × m and B be an n × n matrix. Study the

conditions on A and B such that

(A ⊗ In + Im ⊗ B)v = 0 · 0 = 0

where v 6= 0, i.e. we have an eigenvalue problem with eigenvalue 0. As a

consequence we have

det(A ⊗ In + Im ⊗ B) = 0.

Note that
tr(A ⊗ In + Im ⊗ B) = ntr(A) + mtr(B).
Study first the case with m = n = 2 and A and B the Pauli spin matrices.

Problem 12. Given the hermitian matrices of the three dipole operators
     
0 1 0 0 −i 0 1 0 0
1  1
L1 = √ 1 0 1  , L2 = √  i 0 −i  , L3 =  0 0 0 
2 0 1 0 2 0 i 0 0 0 −1

and the hermitian matrices of five quadrupole operators

   
0 0 1 0 0 −i
W1 =  0 0 0, W2 =  0 0 0  ,
1 0 0 i 0 0
   
0 1 0 0 −i 0
1 1
V1 = √  1 0 −1  , V2 = √ i 0 i ,
2 0 −1 0 2 0 −i 0
 
1 0 0
1 
Q0 = √ 0 −2 0.
3 0 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 63

Kronecker Product and Tensor Product 63

Show that multiplying these eight hermitian matrices by i we obtain a basis

for the semi-simple Lie algebra su(3). Consider the Hamilton operator
Ĥ = κ0 Q0 ⊗ Q0 + κ1 (V1 ⊗ V1 + V2 ⊗ V2 ) + κ2 (W1 ⊗ W1 + W2 ⊗ W2 ).
Find the eigenvalues and eigenvectors of Ĥ.

Problem 13. Let

   
1 0
w1 = , w2 =
0 1
and
2
X
u= tj1 j2 j3 wj1 ⊗ wj2 ⊗ wj3 = w1 ⊗ w2 ⊗ w2 + w2 ⊗ w1 ⊗ w1
j1 ,j2 ,j3 =1
T
=(0 0 0 1 1 0 0 0)
i.e. t122 = 1, t211 = 1 and all other coefficients are equal to 0. Can the
vector u ∈ C8 be written as the Kronecker product of a vector in C2 and a
vector in C4 ? Consider both cases C2 ⊗ C4 and C4 ⊗ C2 .

Problem 14. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Consider the

Hamilton operator
Ĥ = ~ω1 σ1 ⊗ σ1 + ~ω2 σ2 ⊗ σ2 + ~ω3 σ3 ⊗ σ3
acting in the Hilbert space C4 . Show that

eiĤt/~ (σ1 ⊗ I2 )e−iĤt/~ = (σ1 ⊗ I2 ) cos(ω2 t) cos(ω3 t)

+(I2 ⊗ σ1 ) sin(ω2 t) sin(ω3 t)
−(σ2 ⊗ σ3 ) cos(ω2 t) sin(ω3 t)
+(σ3 ⊗ σ2 ) sin(ω2 t) cos(ω3 t)

eiĤt/~ (σ2 ⊗ I2 )e−iĤt/~ = (σ2 ⊗ I2 ) cos(ω3 t) cos(ω1 t)

+(I2 ⊗ σ2 ) sin(ω3 t) sin(ω1 t)
−(σ3 ⊗ σ1 ) cos(ω3 t) sin(ω1 t)
+(σ1 ⊗ σ3 ) sin(ω3 t) cos(ω1 t)

eiĤt/~ (σ3 ⊗ I2 )e−iĤt/~ = (σ3 ⊗ I2 ) cos(ω1 t) cos(ω2 t)

+(I2 ⊗ σ3 ) sin(ω1 t) sin(ω2 t)
−(σ1 ⊗ σ2 ) cos(ω1 t) sin(ω2 t)
+(σ2 ⊗ σ1 ) sin(ω1 t) cos(ω2 t).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 64

64 Problems and Solutions

Problem 15. Let S1 , S2 , S3 be the spin- 12 matrices

     
1 0 1 1 0 −i 1 1 0
S1 = , S2 = , S3 = .
2 1 0 2 i 0 2 0 −1

Solve the eigenvalue problem for the Hamilton operator

Ĥ = ~ω1 (S1 ⊗ S2 ⊗ S3 + S3 ⊗ S1 ⊗ S2 + S2 ⊗ S3 ⊗ S1 )
+~ω2 (S3 ⊗ I2 ⊗ I2 + I2 ⊗ S3 ⊗ I2 + I2 ⊗ I2 ⊗ S3 ).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 65

Chapter 3

Matrix Properties

3.1 Introduction
For finite-dimensional quantum systems finding the norm, eigenvalues, eigen-
vectors, Schmidt rank and inverse (if it exists) of square matrices is impor-
tant. Let A be an n × n matrix over C. Then we can define the sup-norm
kAk := sup kAxk
kxk=1

where kAxk denotes the Euclidean norm in Cn . The Hilbert-Schmidt norm

of a square matrix A is defined as
kAk := (tr(AA∗ ))1/2
where tr denotes the trace.
Let A be an n × n matrix over C. Then the eigenvalue equation is defined
as
Ax = λx
where λ ∈ C is the eigenvalue and x ∈ Cn with x 6= 0 is a corresponding
eigenvector. It follows that x∗ A∗ = λx∗ .

The most important function in quantum computing is the exponential

function of a square matrix A defined by
∞ k
Aj

X A
exp(A) := = lim I + .
j=0
j! k→∞ k

65
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 66

66 Problems and Solutions

An n × n matrix over C is called normal if

A∗ A = AA∗ .

Let A be an n × n normal matrix over C with eigenvalues λ1 , . . . , λn and

corresponding pairwise orthonormal eigenvectors vj (j = 1, . . . , n). Then
the matrix A can be written as (spectral decomposition)
n
X
A= λj vj vj∗ .
j=1

Note that
vj vj∗ , j = 1, . . . , n
are projection matrices, i.e. (vj vj∗ )(vj vj∗ )
= vj vj∗ , (vj vj∗ )∗ = vj vj∗ and
∗ ∗
vj vj = 1 for j = k and vj vk = 0 for j 6= k.

Consider an n × n matrix A over C and the polynomial

p(λ) = det(A − λIn )

with the characteristic equation

p(λ) = 0.

The Cayley-Hamilton theorem states that substituting the matrix A in the

characteristic polynomial results in the n × n zero matrix, i.e.

p(A) = 0n .

Decompositions of square matrices such as the singular value decomposi-

tion, spectral decomposition, polar decomposition and Schur decomposition
are necessary in quantum computing. Any unitary 2n × 2n matrix U can
be decomposed as
   
U1 0 C S U3 0
U=
0 U2 −S C 0 U4

where U1 , U2 , U3 , U4 are 2n−1 × 2n−1 unitary matrices and C and S are the
2n−1 × 2n−1 diagonal matrices

C = diag(cos(α1 ), cos(α2 ), . . . , cos(α2n /2 )),

S = diag(sin(α1 ), sin(α2 ), . . . , sin(α2n /2 ))

where αj ∈ R. This decomposition is called cosine-sine decomposition.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 67

Matrix Properties 67

3.2 Solved Problems

Problem 1. Consider the hermitian 4 × 4 matrix (Hamilton operator)
~ω
Ĥ = (σ1 ⊗ σ1 − σ2 ⊗ σ2 )
2
where ω is the frequency. Find the norm of Ĥ, i.e.
kĤk := sup kĤxk, x ∈ C4 .
kxk=1

Solution 1. There are two methods to find the norm of Ĥ. In the
first method we use the Lagrange multiplier method where the constraint
kxk = 1 can be written as x21 + x22 + x23 + x24 = 1. Since
0 0 0 1 0 0 0 −1
   
0 0 1 0  0 0 1 0 
σ1 ⊗ σ1 =   , σ2 ⊗ σ2 = 
0 1 0 0 0 1 0 0

1 0 0 0 −1 0 0 0
we have
0 0 0 1
 
0 0 0 0
Ĥ = ~ω  .
0 0 0 0
1 0 0 0
Let x = (x1 , x2 , x3 , x4 )T ∈ C4 . We maximize
f (x) := kĤxk2 − λ(x21 + x22 + x23 + x24 − 1)
where λ is the Lagrange multiplier. To find the extrema we solve the four
equations
∂f
= 2~2 ω 2 x1 − 2λx1 = 0
∂x1
∂f
= −2λx2 = 0
∂x2
∂f
= −2λx3 = 0
∂x3
∂f
= 2~2 ω 2 x4 − 2λx4 = 0
∂x4
together with the constraint x21 + x22 + x23 + x24 = 1. The four equations can
be written in the matrix form
 2 2
~ ω −λ 0 0 0 x1 0
   
0 −λ 0 0   x2   0 
  =  .

0 0 −λ 0 x3 0

0 0 0 ~2 ω 2 − λ x4 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 68

68 Problems and Solutions

If λ = 0 then x1 = x4 = 0 and kĤxk = 0, which is a minimum. If λ 6= 0

then x2 = x3 = 0 and x21 + x24 = 1 so that kĤxk = ~ω, which is the
maximum. Thus we find kĤk = ~ω. In the second method we calculate
the positive definite matrix Ĥ ∗ Ĥ and find the square root of the largest
eigenvalue of Ĥ ∗ Ĥ. Since Ĥ ∗ = Ĥ we find the positive semi-definite

1 0 0 0
 
∗ 2 20 0 0 0
Ĥ Ĥ = ~ ω  .
0 0 0 0
0 0 0 1

Thus the maximum eigenvalue is ~2 ω 2 (twice degenerate) and kĤk = ~ω.

Problem 2. Let Ĥ be a hermitian n×n matrix (Hamilton operator) with

eigenvalues E0 , E1 , . . . , En−1 with corresponding normalized eigenvectors
|ψ0 i, ψ1 i, . . . , |ψn−1 i. The quantum correlation function of two n × n
hermitian matrices A and B is given by

1
Qk (t) := hψk |(A(t)B − AB(t) + BA(t) − B(t)A)|ψk i, k = 0, 1, . . . , n − 1
2
where
A(t) = eiĤt/~ Ae−iĤt/~ , B(t) = eiĤt/~ Be−iĤt/~ .

Note that Qk (t) is real valued. Find Qk (t) using the properties

n−1
X
eiĤt/~ |ψk i = eiEk t/~ |ψk i, |ψj ihψj | = In .
j=0

Solution 2. Since
n−1
X
hψk |A(t)B|ψk i = ei(Ek −Ej )t/~ hψk |A|ψj ihψj |B|ψk i
j=0
n−1
X
hψk |AB(t)|ψk i = ei(Ej −Ek )t/~ hψk |A|ψj ihψj |B|ψk i
j=0
n−1
X
hψk |BA(t)|ψk i = ei(Ej −Ek )t/~ hψk |B|ψj ihψj |A|ψk i
j=0
n−1
X
hψk |B(t)A|ψk i = ei(Ek −Ej )t/~ hψk |B|ψj ihψj |A|ψk i
j=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 69

Matrix Properties 69

and utilizing the identity eiα − e−iα ≡ 2i sin(α) yields

n−1
X
Qk (t) = i sin((Ej −Ek )t/~) (hψj |A|ψk ihψk |B|ψj i − hψk |A|ψj ihψj |B|ψk i) .
j=0

Problem 3. (i) Let A and B be two n × n matrices over C. If there

exists a non-singular n × n matrix X such that A = XBX −1 , then A and
B are said to be similar matrices. Show that the spectra (eigenvalues) of
two similar matrices are equal.
(ii) Let A and B be n × n matrices over C. Show that the matrices AB
and BA have the same set of eigenvalues.

Solution 3. (i) We have

det(A − λIn ) = det(XBX −1 − XλIn X −1 ) = det(X(B − λIn )X −1 )
= det(X)det(B − λIn )det(X −1 ) = det(B − λIn ).
(ii) Consider first the case that A is invertible. Then we have
AB = A(BA)A−1 .
Thus AB and BA are similar and therefore have the same set of eigenvalues.
If A is singular we apply the continuity argument: Consider the matrix
A + In . We choose δ > 0 such that A + In is invertible for all , 0 <  < δ.
Thus (A + In )B and B(A + In ) have the same set of eigenvalues for every
 ∈ (0, δ). We equate their characteristic polynomials to obtain
det(λIn − (A + In )B) = det(λIn − B(A + In )), 0 <  < δ.
Since both sides are analytic functions of  we find by letting  → 0+ that
det(λIn − AB) = det(λIn − BA).

Problem 4. Consider a square non-singular matrix A over C. The polar

decomposition theorem states that A can be written as
A = UP
where U is a unitary matrix and P is a positive definite matrix. Thus P is
hermitian. Show that A has a unique polar decomposition.

Solution 4. Since A is invertible, so are A∗ and A∗ A. The positive square

root P of A∗ A is also invertible. Set U := AP −1 . Then U is invertible and
U ∗ U = P −1 A∗ AP −1 = P −1 P 2 P −1 = I
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 70

70 Problems and Solutions

so that U is unitary. Since P is invertible, it is obvious that AP −1 is the

only possible choice for U .

Problem 5. Let A and B be n × n hermitian matrices. Suppose that

A2 = In , B 2 = In (1)

and
[A, B]+ ≡ AB + BA = 0n (2)
where 0n is the n × n zero matrix. Let x ∈ Cn be normalized, i.e. kxk = 1.
Here x is considered as a column vector.
(i) Show that
(x∗ Ax)2 + (x∗ Bx)2 ≤ 1. (3)
(ii) Give an example for the matrices A and B.

Solution 5. (i) Let a, b ∈ R and let r2 := a2 + b2 . The matrix

C = aA + bB

is again hermitian. Then

C 2 = a2 A2 + abAB + baBA + b2 B 2 .

Using the properties (1) and (2) we find

C 2 = a2 In + b2 In = r2 In .

Therefore (x∗ C 2 x) = r2 and −r ≤ a(x∗ Ax) + b(x∗ Bx) ≤ r. Let

a = x∗ Ax, b = x∗ Bx

then a2 + b2 ≤ r or r2 ≤ r. This implies r ≤ 1 and r2 ≤ 1 from which (3)

follows.
(ii) An example is A = σ1 and B = σ2 since σ12 = I2 , σ22 = I2 and
σ1 σ2 + σ2 σ1 = 02 .

Problem 6. Let K be an n × n skew-hermitian matrix K = −K ∗ with

eigenvalues µ1 , . . . , µn (counted according to multiplicity) and the corre-
sponding normalized eigenvectors u1 , . . . , un , where u∗j uk = 0 for k 6= j.
Then K can be written as
n
X
K= µj uj u∗j
j=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 71

Matrix Properties 71

and uj u∗j uk u∗k = 0 for k 6= j and j, k = 1, 2, . . . , n. Note that the n × n

matrices uj u∗j are projection matrices and
n
X
uj u∗j = In .
j=1

(i) Calculate exp(K).

(ii) Every n × n unitary matrix can be written as U = exp(K), where K is
a skew-hermitian matrix. Find U from a given K.
(iii) Use the result from (ii) to find for a given U a possible K.
(iv) Apply the result from (ii) and (iii) to the unitary 2 × 2 matrix
 
cos(θ) sin(θ)
U (θ) = .
− sin(θ) cos(θ)

(v) Apply the result from (ii) and (iii) to the 2 × 2 unitary matrix
 
cos(θ) −eiφ sin(θ)
V (θ, φ) = .
e−iφ sin(θ) cos(θ)

(vi) Every hermitian matrix H can be written as H = iK, where K is a

skew-hermitian matrix. Find H for the examples given above.

Solution 6. (i) Using the properties of the n × n matrix uj u∗j we find

 
Xn n
X
exp(K) = exp  µj uj u∗j  = eµj uj u∗j .
j=1 j=1

(ii) From U = exp(K) we find

n
X
U= eµj uj u∗j
j=1

where uj (j = 1, 2, . . . , n) are the normalized eigenvectors of U .

(iii) The matrix K is given by
n
X
K= ln(λj )uj u∗j
j=1

where λj (j = 1, 2, . . . , n) are the eigenvalues of U and uj are the normalized

eigenvectors of U . Note that the eigenvalues of U are of the form exp(iα)
with α ∈ R. Thus we have ln(eiα ) = iα.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 72

72 Problems and Solutions

(iv) The eigenvalues of the matrix U (θ) are eiθ and e−iθ with the corre-
sponding normalized eigenvectors
   
1 1 1 1
u1 = √ , u2 = √ .
2 i 2 −i
Thus
     
iθ 1 −i iθ 1 i 0 θ
K(θ) = ln(e iθ
)u1 u∗1 +ln(eiθ )u2 u2 = − = .
2 i 1 2 −i 1 −θ 0

(v) For the matrix V (θ, φ) the eigenvalues are e−iθ and eiθ with the corre-
sponding normalized eigenvectors
   
1 1 1 1
√ −iφ , √ −iφ .
2 ie 2 −ie
Thus
 
0 −θeiφ
K(θ, φ) = ln(e−iθ )u1 u∗1 + ln(eiθ )u2 u∗2 = .
θe−iφ 0

(vi) For U (θ) we find the matrix

 
0 1
iθ .
−1 0

For V (θ, φ) we find the matrix

 
0 −iθeiφ
.
iθe−iφ 0

Problem 7. Let A and B be n × n hermitian matrices. Suppose that

A2 = A, B2 = B (1)

and
[A, B]+ ≡ AB + BA = 0n (2)
n
where 0n is the n × n zero matrix. Let x ∈ C be normalized, i.e. kxk = 1.
Here x is considered as a column vector. Show that

(x∗ Ax)2 + (x∗ Bx)2 ≤ 1. (3)

Solution 7. For an arbitrary n × n hermitian matrix M we have

0 ≤ (x∗ (M − (x∗ M x)In )2 x) = (x∗ (M 2 − 2(x∗ M x)M + (x∗ M x)2 In )x)

= (x∗ M 2 x) − 2(x∗ M x)2 + (x∗ M x)2 = (x∗ M 2 x) − (x∗ M x)2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 73

Matrix Properties 73

Thus
0 ≤ (x∗ M 2 x) − (x∗ M x)2 or (x∗ M x)2 ≤ (x∗ M 2 x).
Thus for A = M we have using (1)

(x∗ Ax)2 ≤ x∗ Ax

and therefore 0 ≤ (x∗ Ax) ≤ 1. Similarly 0 ≤ (x∗ Bx) ≤ 1. Let a, b ∈ R,

r2 := a2 + b2 and C := aA + bB. Then

C 2 = a2 A2 + b2 B 2 + abAB + baBA.

Using (1) and (2) we arrive at C 2 = a2 A + b2 B. Thus

(x∗ Cx)2 ≤ (x∗ C 2 x) ≤ a2 + b2 .

Let a := (x∗ Ax), b := (x∗ Bx) then (x∗ Cx) = a2 + b2 = r2 and therefore
(r2 )2 ≤ r2 which implies that r2 ≤ 1 and thus (3) follows.

Problem 8. Let A, B be n × n matrices over C. Assume that

[A, [A, B]] = [B, [A, B]] = 0n . (1)

Show that
1
eA+B = eA eB e− 2 [A,B] (2a)
1
eA+B = eB eA e+ 2 [A,B] . (2b)
Hint. Use the technique of parameter differentiation. Consider the matrix-
valued function
f () = eA eB
where  is a real parameter and calculate the derivative df /d.

Solution 8. If we differentiate f () with respect to  we find

df
= AeA eB + eA eB B = (A + eA Be−A )f ()
d

since eA e−A = In . Owing to (1) we have

eA B−A = B + [A, B].

Thus we obtain the linear matrix-valued differential equation

df
= ((A + B) + [A, B])f ().
d
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 74

74 Problems and Solutions

Since the matrix A + B commutes with [A, B] we may treat A + B and

[A, B] as ordinary commuting variables and integrate this linear differential
equation with the initial conditions f (0) = In . We find
2 2
f () = e(A+B)+( /2)[A,B]
= e(A+B) e( /2)[A,B]

since A + B commutes with [A, B]. If we set  = 1 and multiply both sides
by e−[A,B]/2 then (2a) follows. Likewise we can prove the second form of
the identity (2b).

Problem 9. Let A be an n × n matrix. Assume that the inverse matrix

of A exists. The inverse matrix can be calculated as follows (Csanky’s
algorithm). Let
p(x) := det(xIn − A) (1)
where In is the n × n unit matrix. The roots are, by definition, the eigen-
values λ1 , λ2 , . . . , λn of A. We write

p(x) = xn + c1 xn−1 + · · · + cn−1 x + cn (2)

where cn = (−1)n det(A). Since A is nonsingular we have cn 6= 0 and vice

versa. The Cayley-Hamilton theorem states that

p(A) = An + c1 An−1 + · · · + cn−1 A + cn In = 0n . (3)

Multiplying this equation with A−1 we obtain

1
A−1 = (An−1 + c1 An−2 + · · · + cn−1 In ). (4)
−cn
If we have the coefficients cj we can calculate the inverse matrix A. Let
n
X
sk := λkj .
j=1

Then the sj and cj satisfy the following n × n lower triangular system of

linear equations

1 0 0 ... 0 c1 −s1
    
 s1 2 0 . . . 0   c2   −s2 
 s2 s1 3 . . . 0   c3  =  −s3  .
    
 . .. .. . . .   .   . 
 .. . . . ..   ..   .. 
sn−1 sn−2 . . . s1 n cn −sn

Since
tr(Ak ) = λk1 + λk2 + · · · + λkn = sk
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 75

Matrix Properties 75

we find sk for k = 1, 2, . . . , n. Thus we can solve the linear equation for

cj . Finally, using (4) we obtain the inverse matrix of A. Apply Csanky’s
algorithm to the 4 × 4 permutation matrix

0 1 0 0
 
0 0 1 0
U = .
0 0 0 1
1 0 0 0

Solution 9. Since
0 0 1 0 0 0 0 1
   
0 0 0 1 1 0 0 0
U2 =  , U3 = 
 
1 0 0 0 0 1 0 0

0 1 0 0 0 0 1 0

and U 4 = I4 we find

tr(U ) = 0 = s1 , tr(U 2 ) = 0 = s2 , tr(U 3 ) = 0 = s3 , tr(U 4 ) = 4 = s4 .

We obtain the system of linear equations

1 0 0 0 c1 0
    
0 2 0 0   c2   0 
  = 
0 0 3 0 c3 0
 
0 0 0 4 c4 −4

with the solution c1 = 0, c2 = 0, c3 = 0, c4 = −1. Thus the inverse matrix

of U is given by
0 0 0 1
 
1 0 0 0
U −1 = U 3 =  .
0 1 0 0
0 0 1 0

Problem 10. Let

     
+ 0 1 − 0 0 1 1 0
J := , J := , J3 := .
0 0 1 0 2 0 −1
−
+ +
+J − )
(i) Let  ∈ R. Find the matrices eJ , eJ , e(J .
(ii) Let r ∈ R. Show that
+
+J − ) −
tanh(r) 2J3 ln(cosh(r)) J + tanh(r)
er(J ≡ eJ e e .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 76

76 Problems and Solutions

Solution 10. (i) Using the expansion for an n × n matrix A

∞ j j
X  A
exp(A) =
j=0
j!

we find
       
+ 1 0 0 1 − 1 0 0 0
eJ = + , eJ = +
0 1 0 0 0 1 1 0

and    
(J + +J − ) 1 0 0 1
e = cosh() + sinh().
0 1 1 0
(ii) Since  
cosh(r) 0
e2J3 ln(cosh(r)) = ,
0 1/ cosh(r)
1/ cosh(r) + tanh(r) sinh(r) ≡ cosh(r) and using the results from (i) we find
the identity.

Problem 11. The Heisenberg commutation relation can be written as

[p̂, q̂] = −i~I

where p̂ := −i~∂/∂q and I is the identity operator. Let α, β ∈ R and

U (α) = exp(iαp̂), V (β) = exp(iβ q̂).

Then using the Baker-Campbell-Hausdorff formula we find

U (α)V (β) = exp(iαβ)V (β)U (α).

This is called the Weyl representation of Heisenberg’s commutation relation.

Can we find finite-dimensional n × n unitary matrices U (U 6= In ) and V
(V 6= In ) such that
U V = ωV U
with ω ∈ C, ω n = 1 ?

Solution 11. Such matrices can be found, namely the permutation matrix

0 1 0 ... 0
 
0 0 1 ... 0
. . . .
. . . . . ... 

U := . . . 
0 0 0 ... 1
1 0 0 ... 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 77

Matrix Properties 77

and the diagonal matrix

1 0 0 ... 0
 
0 ω 0 ... 0 
0 0 ω2 ... 0
 
V := 
.
.
 .. .. .. .. .. 
. . . . 
0 0 0 ... ω n−1

Problem 12. Let U be the n × n unitary matrix

0 1 0 ... 0
 
0 0 1 ... 0
. . . .. .. 
U :=  . . . .
. . . .
0 0 0 ... 1
1 0 0 ... 0

and V be the n × n unitary diagonal matrix (ω ∈ C)

1 0 0 ... 0
 
0 ω 0 ... 0 
0 0 ω2 ... 0
 
V := 
.

 .. .. .. .. .. 
. . . . 
0 0 0 ... ω n−1

where ω n = 1 (ω 6= 1). Then the set of matrices

{ U j V k : j, k = 0, 1, 2, . . . , n − 1 }

provide a basis in the Hilbert space for all n × n matrices with the scalar
product
1
hA, Bi := tr(AB ∗ )
n
for n × n matrices A and B. Write down the basis for n = 2.

Solution 12. For n = 2 we have the combinations

(j, k) ∈ { (0, 0), (0, 1), (1, 0), (1, 1) }.

This yields the orthonormal basis (where ω = −1)

       
1 0 0 1 1 0 0 −1
I2 = , σ1 = , σ3 = , −iσ2 = .
0 1 1 0 0 −1 1 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 78

78 Problems and Solutions

Problem 13. An n × n circulant matrix C is given by

 c c1 c2 ... c 
0 n−1
 cn−1 c0 c1 . . . cn−2 
cn−2 cn−1 c0 . . . cn−3  .
 
C := 
 .
 . .. .. .. .. 
. . . . .

c1 c2 c3 ... c0
For example, the permutation matrix
0 1 0 ... 0
 
0 0 1 ... 0
. . .. .. .. 
P :=  . . .
. . . .
0 0 0 ... 1
1 0 0 ... 0
is a circulant matrix. It is also called the n×n primary permutation matrix.
(i) Let C and P be the matrices given above. Let
f (λ) = c0 + c1 λ + · · · + cn−1 λn−1 .
Show that C = f (P ).
(ii) Show that C is a normal matrix, that is C ∗ C = CC ∗ .
(iii) Show that the eigenvalues of C are f (ω k ), k = 0, 1, . . . , n − 1, where ω
is the nth primitive root of unity.
(iv) Show that
det(C) = f (ω 0 )f (ω 1 ) · · · f (ω n−1 ).
(v) Show that F ∗ CF is a diagonal matrix, where F is the unitary matrix
with (j, k)-entry equal to
1
√ ω (j−1)(k−1) , j, k = 1, . . . , n.
n

Solution 13. (i) Direct calculation of

f (P ) = c0 In + c1 P + c2 P 2 + · · · + cn−1 P n−1
yields the matrix C, where In is the n × n unit matrix. Notice that P 2 ,
P 3 , . . . , P n−1 are permutation matrices.
(ii) We have P P ∗ = P ∗ P . If two n × n matrices A and B commute, then
g(A) and h(B) commute, where g and h are polynomials. Thus C is a
normal matrix.
(iii) The characteristic polynomial of P is
n−1
Y
det(λIn − P ) = λn − 1 = (λ − ω k ).
k=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 79

Matrix Properties 79

Thus the eigenvalues of P and P j are, respectively, ω k and ω jk , where

k = 0, 1, . . . , n − 1. It follows that the eigenvalues of C = f (P ) are f (ω k ),
k = 0, 1, . . . , n − 1.
(iv) Using the result from (iii) we find
n−1
Y
det(C) = f (ω k ).
k=0

(v) For each k = 0, 1, . . . , n − 1, let

xk = (1, ω k , ω 2k , . . . , ω (n−1)k )T
T
where denotes the transpose. If follows that

P xk = (ω k , ω 2k , . . . , ω (n−1)k , 1)T = ω k xk

and
Cxk = f (P )xk = f (ω k )xk .
Thus the vectors xk are the eigenvectors of P and C corresponding to the
respective eigenvalues ω k and f (ω k ), k = 0, 1, . . . , n − 1. Since
n−1 n−1 
X X 0 j 6= k
hxj , xk i ≡ x∗j xk = ω k` ω j` = ω (j−k)` =
n j=k
`=0 `=0

we find that  
1 1 1
√ x0 , √ x1 , . . . , √ xn−1
n n n
is an orthonormal basis in the Hilbert space Cn . Thus we obtain the unitary
matrix
1 1 1 ... 1
 
1 2 n−1
ω ω ... ω 
1  1 ω 2
ω 4
. . . ω 2(n−1) 
F =√   
n  .. .. .. .. .. 
. . . . . 
n−1 2(n−1) (n−1)(n−1)
1 ω ω ... ω
such that
F ∗ CF = diag(f (ω 0 ), f (ω 1 ), . . . , f (ω n−1 )).
The matrix F is unitary and is called the Fourier matrix.

Problem 14. An n × n matrix A is called a Hadamard matrix if each

entry of A is 1 or −1 and if the rows or columns of A are orthogonal, i.e.

AAT = nIn or AT A = nIn .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 80

80 Problems and Solutions

Note that AAT = nIn and AT A = nIn are equivalent. Hadamard matrices
Hn of order 2n can be generated recursively by defining
   
1 1 Hn−1 Hn−1
H1 = , Hn =
1 −1 Hn−1 −Hn−1

for n ≥ 2. Show that the eigenvalues of Hn are given by +2n/2 and −2n/2
each of multiplicity 2n−1 .

Solution 14. We use induction on n. The case n = 1 is obvious. Now

for n ≥ 2 we have

 
 λI − Hn−1 −Hn−1 
det(λI − Hn ) =  
−Hn−1 λI + Hn−1 
2
= det((λI − Hn−1 )(λI + Hn−1 ) − Hn−1 ).

Thus

det(λI − Hn ) = det(λ2 I − 2Hn−1

2
)
√ √
= det(λI − 2Hn−1 ) det(λI + 2Hn−1 ).
√
This shows that each eigenvalue µ of Hn−1 generates two eigenvalues ± 2µ
of Hn . The assertion then follows by the induction hypothesis, for Hn−1
has eigenvalues +2(n−1)/2 and −2(n−1)/2 each of multiplicity 2n−2 .

Problem 15. Let U be an n × n unitary matrix. Then U can be written

as
U = V diag(λ1 , λ2 , . . . , λn )V ∗
where λ1 , λ2 , . . . , λn are the eigenvalues of U and V is an n × n unitary
matrix. Let  
0 1
U= .
1 0
Find the decomposition for U given above.

Solution 15. The eigenvalues of U are +1 and −1. Thus we have

U = V diag(1, −1)V ∗

with  
1 1 1
V =√ .
2 1 −1
Therefore V = V ∗ . The columns of V are the eigenvectors of U .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 81

Matrix Properties 81

Problem 16. An n × n hermitian matrix A over the complex numbers is

called positive semidefinite (written as A ≥ 0), if

x∗ Ax ≥ 0 for all x ∈ Cn .

Show that for every A ≥ 0, there exists a unique B ≥ 0 so that B 2 = A.

Solution 16. Let A = U ∗ diag(λ1 , . . . , λn )U , where U is unitary. We take

1/2
B = U ∗ diag(λ1 , . . . , λ1/2
n )U.

Then the matrix B is positive semidefinite and B 2 = A since U ∗ U = In .

To show the uniqueness, suppose that C is an n × n positive semidefinite
matrix satisfying C 2 = A. Since the eigenvalues of C are the nonnegative
square roots of the eigenvalues of A, we can write
1/2 ∗
C = V diag(λ1 , . . . , λ1/2
n )V

for some unitary matrix V . Then the identity C 2 = A = B 2 yields

T diag(λ1 , . . . , λn ) = diag(λ1 , . . . , λn )T

where T = U V . This yields tjk λk = λj tjk . Thus

1/2 1/2
tjk λk = λj tjk .

Hence
1/2 1/2
T diag(λ1 , . . . , λ1/2 1/2
n ) = diag(λ1 , . . . , λn )T.

Since T = U V it follows that B = C.

Problem 17. An n × n matrix A over the complex numbers is said to

be normal if it commutes with its conjugate transpose A∗ A = AA∗ . The
matrix A can be written
Xn
A= λj Ej
j=1

where λj ∈ C are the eigenvalues of A and Ej are n × n matrices satisfying

n
X
Ej2 = Ej = Ej∗ , Ej Ek = 0n if j 6= k, Ej = In .
j=1

Let n = 2. Consider the NOT gate

 
0 1
A= .
1 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 82

82 Problems and Solutions

Find the decomposition of A given above.

Solution 17. The eigenvalues of A are given by λ1 = +1, λ2 = −1. The

matrices Ej are constructed from the normalized eigenvectors of A. The
normalized eigenvectors of A are given by
   
1 1 1 1
v1 = √ , v2 = √ .
2 1 2 −1
Thus
   
1 1 1 1 1 −1
E1 = v1 v1∗ = , E2 = v2 v2∗ = .
2 1 1 2 −1 1

Problem 18. Let X be an n × n matrix over C with X 2 = In , where In

is the n × n identity matrix. Let z ∈ C.
(i) Show that ezX = In cosh(z) + X sinh(z).
(ii) Let z = iα, where α ∈ R. Simplify the result from (i).

Solution 18. (i) We have

∞ ∞
X (zX)k X zk X k 1 2 2 1
ezX := = = In + zX + z X + z3X 3 + · · · .
k! k! 2! 3!
k=0 k=0

Since X 2 = In we have
   
1 1 1 1
ezX = In 1 + z 2 + z 4 + · · · + X z + z 3 + z 5 + · · ·
2! 4! 3! 5!
= In cosh(z) + X sinh(z).

(ii) Since cosh(iα) = cos(α), sinh(iα) = i sin(α) we obtain

eiαX = In cos(α) + iX sin(α).

Problem 19. Let σ1 , σ2 be the Pauli spin matrices.

(i) Consider

V = exp(i(π/4)σ1 ), W = exp(i(π/4)σ2 ).

Show that    
1 1 i 1 1 1
V =√ , W =√ .
2 i 1 2 −1 1
(ii) Let
V = exp(i(π/4)σ1 ), W = exp(i(π/4)σ2 )
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 83

Matrix Properties 83

be two unitary 2 × 2 matrices. Calculate the matrices V ∗ σ3 V , W ∗ σ3 W .

Solution 19. (i) Using σ12 = I2 , σ22 = I2 we find

 
1 1 1 i
V = In cos(π/4) + iσ1 sin(π/4) = √ (I2 + iσ1 ) = √
2 2 i 1

and
 
1 1 1 1
W = In cos(π/4) + iσ2 sin(π/4) = √ (I2 + iσ2 ) = √ .
2 2 −1 1

(ii) We have
   
∗ 1 i 1 i
V σ3 V = √ I2 − √ σ1 σ3 √ I2 + √ σ1
2 2 2 2
1 i i 1
= σ3 + σ3 σ1 − σ1 σ3 + σ1 σ3 σ1
2 2 2 2
= −σ2

and
   
1 i 1 i
W ∗ σ3 W = √ I2 − √ σ2 σ3 √ I2 + √ σ2
2 2 2 2
1 i i 1
= σ3 + σ3 σ2 − σ2 σ3 + σ2 σ3 σ2
2 2 2 2
= σ1 .

Problem 20. Find the matrices

e(iπ/4)σ2 σ1 e−(iπ/4)σ2 , e(iπ/4)σ2 σ3 e−(iπ/4)σ2 .

Use the technique of parameter differentiation

f () = eσ2 σ1 e−σ2

with f ( = 0) = σ1 .

Solution 20. Differentiation of f with respect to  and using [σ2 , σ1 ] =

−2iσ3 yields
df
= −2ieσ2 σ3 e−σ2
d
with df ( = 0)/d = −2iσ3 . The second derivative and using [σ2 , σ3 ] = 2iσ1
yields
d2 f
= 4f ().
d2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 84

84 Problems and Solutions

The solution of this second order linear differential equation with constant
coefficients is
f () = C1 cosh(2) + C2 sinh(2).
Inserting the initial values provides

f () = cosh(2)σ1 − i sinh(2)σ3 .

Since  = iπ/4 we have cosh(iπ/2) = cos(π/2) = 0 and sinh(iπ/2) =

i sin(π/2) = i. Thus f ( = iπ/4) = σ3 or

e(iπ/4)σ2 σ1 e−(iπ/4)σ2 = σ3 .

Analogously we find e(iπ/4)σ2 σ3 e−(iπ/4)σ2 = σ1 .

Problem 21. Let

V = exp(i(π/4)σ1 ) ⊗ exp(i(π/4)σ1 ), W = exp(i(π/4)σ2 ) ⊗ exp(i(π/4)σ2 ).

Calculate V ∗ (σ3 ⊗ σ3 )V and W ∗ (σ3 ⊗ σ3 )W .

Solution 21. We have

V ∗ (σ3 ⊗ σ3 )V = (exp(−i(π/4)σ1 )σ3 exp(i(π/4)σ1 ))

⊗(exp(−i(π/4)σ1 )σ3 exp(i(π/4)σ1 ))
= σ2 ⊗ σ2 .

Analogously W ∗ (σ3 ⊗ σ3 )W = σ1 ⊗ σ1 .

Problem 22. Let X be an n × n matrix over C. Assume that X 2 = In .

Let Y be an arbitrary n × n matrix over C. Let z ∈ C.
(i) Calculate exp(zX)Y exp(−zX) using the Baker-Campbell-Hausdorff for-
mula
z2 z3
ezX Y e−zX = Y + z[X, Y ] + [X, [X, Y ]] + [X, [X, [X, Y ]]] + · · · .
2! 3!
(ii) Calculate exp(zX)Y exp(−zX) by first calculating exp(zX) and exp(−zX)
and then doing the matrix multiplication. Compare the two methods.

Solution 22. (i) Using X 2 = In we find for the first three commutators

[X, [X, Y ]] = [X, XY − Y X] = 2(Y − XY X)

[X, [X, [X, Y ]]] = 22 [X, Y ]
[X, [X, [X, [X, Y ]]]] = 23 (Y − XY X).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 85

Matrix Properties 85

If the number of X’s in the commutator is even (say m, m ≥ 2) we have

[X, [X, . . . [X, Y ] . . .] = 2m−1 (Y − XY X).

If the number of X’s in the commutator is odd (say m, m ≥ 3) we have

[X, [X, . . . [X, Y ] . . .] = 2m−1 [X, Y ].

Thus
21 z 2 23 z 4 22 z 3
   
zX −zX
e Ye =Y 1+ + + · · · + [X, Y ] z + + ···
2! 4! 3!
 1 2
23 z 4

2 z
−XY X + + ··· .
2! 4!
Consequently

ezX Y e−zX = Y cosh2 (z) + [X, Y ] sinh(z) cosh(z) − XY X sinh2 (z).

(ii) Using the expansion

∞
X (zX)j
ezX =
j=0
j!

and X 2 = In we have

ezX = In cosh(z) + X sinh(z), e−zX = In cosh(z) − X sinh(z).

Matrix multiplication yields

ezX Y e−zX = Y cosh2 (z) + [X, Y ] sinh(z) cosh(z) − XY X sinh2 (z).

Problem 23. The definition of the Lie group SU (2) is

SU (2) := { A : A a 2 × 2 complex matrix, det(A) = 1, AA∗ = A∗ A = I2 } .

In the name SU (2), the S stands for special and refers to the condition
det(A) = 1 and the U stands for unitary and refers to the conditions AA∗ =
A∗ A = I2 . Show that SU (2) can also be defined as

SU (2) := { x0 I2 + ixT σ : (x0 , x)T ∈ R4 , x20 + kxk2 = 1 }

where σ = (σ1 , σ2 , σ3 )T are the Pauli spin matrices. Here xT σ := x1 σ1 +

x2 σ2 + x3 σ3 .

Solution 23. Let A be any 2 × 2 complex matrix. Then A can be written

as
A = a0 I2 + iaT σ
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 86

86 Problems and Solutions

with a = (a1 , a2 , a3 )T ∈ C3 . Thus we have

AA∗ = (a0 I2 + iaT σ)(a0 I2 − iaT σ)

= |a0 |2 I2 + ia0 aT σ − ia0 aT σ + aT aI2 + i(a × a)T σ
= (|a0 |2 + kak2 )I2 + i(a0 a − a0 a + a × a)T σ

where we used that σj∗ = σj for j = 1, 2, 3. It follows that

AA∗ = I2 ⇔ |a0 |2 + kak2 = 1, a0 a − a0 a + a × a = 0.

First, suppose that a 6= 0. Since a × a is orthogonal to both a and a, the

equation a0 a − a0 a + a × a = 0 can only be satisfied if a × a = 0. That is,
only if a and a are parallel. Since a and a have the same length, this is the
case only if a = e−2iθ a for some real number θ. This can be written as

e−iθ a = e−iθ a

which says that x = e−iθ a is real. Substituting a = eiθ x into a0 a − a0 a +

a × a = 0 gives
eiθ a0 x − e−iθ a0 x = 0.
This forces a0 = eiθ x0 for some real x0 . If a = 0, we may still choose θ so
that a0 = eiθ x0 . We have shown that

AA∗ = I2 ⇔ A = eiθ (x0 I2 + ixT σ)

for some (x0 , x)T ∈ R4 with |x0 |2 + kxk2 = 1 and some θ ∈ R. Since

det(A) = det(eiθ (x0 I2 + ixT · σ))

= e2iθ (x20 + x21 + x22 + x23 )
= e2iθ

we have that det(A) = 1 if and only if eiθ = ±1. If eiθ = −1, we can absorb
the −1 into the vector (x0 , x)T .

Problem 24. Let A, B be two n × n matrices over C. We introduce the

scalar product
tr(AB ∗ ) 1
hA, Bi := = tr(AB ∗ ).
tr(In ) n
The Lie group SU (N ) is defined by the complex n × n matrices U

SU (N ) := { U : U ∗ U = U U ∗ = In , det(U ) = 1 }.

The dimension is N 2 − 1. The semi-simple Lie algebra su(N ) is defined by

the n × n matrices X

su(N ) := { X : X ∗ = −X , tr(X) = 0 }.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 87

Matrix Properties 87

(i) Let U ∈ SU (N ). Calculate the scalar product hU, U i.

(ii) Let A be an arbitrary complex n×n matrix. Let U ∈ SU (N ). Calculate
the scalar product hU A, U Ai.
(iii) Consider the Lie algebra su(2). Provide a basis. The elements of the
basis should be orthogonal to each other with respect to the scalar product
given above. Calculate the commutator of these matrices.

Solution 24. (i) We have

1
hU, U i = tr(U U ∗ ) = 1
n
where we used that U ∗ = U −1 .
(ii) We obtain
1 1 1
hU A, U Ai = tr(U A(U A)∗ ) = tr(U AA∗ U ∗ ) = tr(AA∗ ) = hA, Ai
n n n
where we used the cyclic invariance of the trace.
(iii) We are looking for three linear independent 2 × 2 matrices which are
traceless and skew-hermitian matrices. A choice is
     
i 0 0 −1 0 i
τ1 = , τ2 = , τ3 = .
0 −i 1 0 i 0

The matrices are also orthogonal to each other using the scalar product
given above. For the commutators we find

[τ1 , τ2 ] = −2τ3 , [τ3 , τ1 ] = −2τ2 , [τ2 , τ3 ] = −2τ1 .

Problem 25. Let H be an n × n matrix which depends on n real pa-

rameters 1 , 2 ,. . . , n , where we assume that we can differentiate H with
respect to all ’s. Let β > 0 and Z(β) := tr(exp(−βH)).
(i) Show that
Z β
∂ −βH ∂H −τ H
e ≡− dτ e(τ −β)H e . (1)
∂j 0 ∂j
(ii) Show that  
∂ ∂
Z = −βtr He−βH .
∂j ∂j

Solution 25. (i) We set

∂ −βH
f (β, 1 , . . . , n ) := eβH e .
∂j
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 88

88 Problems and Solutions

It follows that
 
∂f ∂ −βH ∂ ∂ −βH
= HeβH e + eβH e .
∂β ∂j ∂β ∂j
Since
   
∂ ∂ −βH ∂ ∂ −βH ∂
e = e =− He−βH
∂β ∂j ∂j ∂β ∂j
∂H −βH ∂ −βH
=− e −H e
∂j ∂j
we obtain
∂f ∂H −βH
= −eβH e
∂β ∂j
with the initial value f (0, 1 , . . . , n ) = 0. Integrating provides
Z β
∂H −τ H
f (β, 1 , . . . , n ) = − eτ H e dτ.
0 ∂j

Multiplying by exp(−βH) yields identity (1).

(ii) We have
 
∂Z ∂ ∂
= tr(exp(−βH)) = tr exp(−βH)
∂j ∂j ∂j
Z β !
(τ −β)H ∂H −τ H
= −tr dτ e e
0 ∂j
 
−βH ∂H
= −βtr e
∂j

where we used the result from (i) and the cyclic invariance of the trace.

Problem 26. To calculate exp(A) we can also use the Cayley-Hamilton

theorem and the Putzer method. Using the Cayley-Hamilton theorem we
can write

f (A) = an−1 An−1 + an−2 An−2 + · · · + a2 A2 + a1 A + a0 In (1)

where the complex numbers a0 , a1 , . . . , an−1 are determined as follows:

Let
r(λ) := an−1 λn−1 + an−2 λn−2 + · · · + a2 λ2 + a1 λ + a0
which is the right-hand side of (1) with Aj replaced by λj (j = 0, 1, . . . , n −
1). For each distinct eigenvalue λj of the matrix A, we consider the equation

f (λj ) = r(λj ). (2)

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 89

Matrix Properties 89

If λj is an eigenvalue of multiplicity k, for k > 1, then we consider also the

following equations
 
f 0 (λ)|λ=λj = r0 (λ)|λ=λj , · · · , f (k−1) (λ) = r(k−1) (λ) .
 
λ=λj λ=λj

Any unitary matrix U can be written as U = exp(iK), where K is hermi-

tian. Apply this method to find K for the Hadamard gate
 
1 1 1
UH = √ .
2 1 −1

Solution 26. The hermitian 2 × 2 matrix K is given by

 
a b
K= , a, c ∈ R, b ∈ C.
b c
Then we find the condition on a, b and c such that eiK = UH . The eigen-
values of iK are given by
i(a + c) 1 p
λ1,2 = ± 2ac − a2 − c2 − 4bb.
2 2
We set in the following
p
∆ := λ1 − λ2 = 2ac − a2 − c2 − 4bb.
To apply the method given above we have
r(λ) = α1 λ + α0 = f (λ) = eλ .
Thus we obtain the two equations eλ1 = α1 λ1 + α0 , eλ2 = α1 λ2 + α0 . It
follows that
eλ1 − eλ2 eλ2 λ1 − eλ1 λ2
α1 = , α0 = .
λ1 − λ2 λ1 − λ2
We have the condition
   
iK iα1 a + α0 iα1 b 1 1 1
e = α1 iK + α0 I2 = =√ .
iα1 b iα1 c + α0 2 1 −1
We obtain the four equations
1 1 1 1
iα1 a + α0 = √ , iα1 c + α0 = − √ , iα1 b = √ , iα1 b = √ .
2 2 2 2
From the last two equations we find that b = b, i.e. b is real. From the first
two equations we find α0 = −iα1 (a + c)/2 and therefore, using the last two
equations, c = a − 2b. Thus
   
iα1 a + α0 iα1 b iα1 b iα1 b
= .
iα1 b iα1 c + α0 iα1 b −iα1 b
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 90

90 Problems and Solutions

From the eigenvalues of eiK we find eλ1 − eλ2 = 2 and

p √
∆ = 2ac − a2 − c2 − 4b2 = 2 2ib.
√ √
Furthermore λ1 = i(a − b) + 2ib, λ2 = i(a − b) − 2ib. Thus we arrive at
the equation √ √
ei(a−b)+ 2ib − ei(a−b)− 2ib = 2.
√
It follows that iei(a−b) sin( 2b) = 1 and therefore
√ √
i cos(a − b) sin( 2b) − sin(a − b) sin( 2b) = 1

with a solution
   
π π 1 π 1
b= √ , a= 3+ √ , c = a − 2b = 3− √ .
2 2 2 2 2 2
Then the matrix K is given by
 √ √     
π 3 + 1/√ 2 1/ 2√ 3π 1 0 π 1 1 1
K= = + ·√ .
2 1/ 2 3 − 1/ 2 2 0 1 2 2 1 −1

We note that the second matrix on the right-hand side is the Hadamard
gate again.

Problem 27. Let A, B, C2 , ..., Cm , ... be n × n matrices over C. The

Zassenhaus formula is given by

exp(A + B) = exp(A) exp(B) exp(C2 ) · · · exp(Cm ) · · ·

The left-hand side is called the disentangled form and the right hand side
is called the undisentangled form. Find C2 , C3 , . . . , using the comparison
method. In the comparison method the disentangled and undisentangled
form are expanded in terms of an ordering scalar α and matrix coefficients
of equal powers of α are compared. From

exp(α(A + B)) = exp(αA) exp(αB) exp(α2 C2 ) exp(α3 C3 ) · · ·

we obtain
∞ ∞
X αk X αr0 +r1 +2r2 +3r3 +... r0 r1 r2 r3
(A + B)k= A B C2 C3 · · ·
k! r0 ,r1 ,r2 ,r3 ,...=0
r0 !r1 !r2 !r3 ! · · ·
k=0

(i) Find the matrices C2 and C3 .

(ii) Assume that [A, [A, B]] = 0n and [B, [A, B]] = 0n . What conclusion
can we draw for the Zassenhaus formula?
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 91

Matrix Properties 91

Solution 27. (i) For α2 we have the decompositions (r0 , r1 , r2 ) = (2, 0, 0),
(1, 1, 0), (0, 2, 0), (0, 0, 1). Thus we obtain

(A + B)2 = A2 + 2AB + B 2 + 2C2 .

Thus it follows that

1
C2 = − [A, B].
2
For α3 we obtain

(A + B)3 = A3 + 3A2 B + 3AB 2 + B 3 + 6AC2 + 6BC2 + 6C3 .

Using C2 given above we obtain

1 1
C3 = [B, [A, B]] + [A, [A, B]].
3 6
(ii) Since [B, [A, B]] = 0n and [A, [A, B]] = 0n we find that C3 = C4 =
· · · = 0. Thus

exp(α(A + B)) = exp(αA) exp(αB) exp(−α2 [A, B]/2).

Problem 28. Let H be a hermitian n × n matrix. Show that exp(H) is

a positive definite matrix.

Solution 28. If H is hermitian then H 2 , H 3 etc are hermitian and also

exp(H). Let λj (j = 1, 2, . . . , n) be the real eigenvalues of H since H is
hermitian. Then eλj (j = 1, 2, . . . , n) are the real eigenvalues of exp(H) and
obviously eλj > 0 for (j = 1, 2, . . . , n). Thus exp(H) is a positive definite
matrix.

Problem 29. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Does the set of
4 × 4 matrices { I2 ⊗ I2 , σ1 ⊗ σ1 , −σ2 ⊗ σ2 , σ3 ⊗ σ3 } form a group under
matrix multiplication?

Solution 29. We have

σ1 σ2 = iσ3 , σ2 σ1 = −iσ3 , σ2 σ3 = iσ1 ,

σ3 σ2 = iσ1 , σ3 σ1 = iσ2 , σ1 σ3 = −iσ2 .

Thus

(σ1 ⊗ σ1 )(−σ2 ⊗ σ2 ) = −(σ1 σ2 ) ⊗ (σ1 σ2 ) = σ3 ⊗ σ3

(−σ2 ⊗ σ2 )(σ3 ⊗ σ3 ) = −(σ2 σ3 ) ⊗ (σ2 σ3 ) = σ1 ⊗ σ1
(σ3 ⊗ σ3 )(σ1 ⊗ σ1 ) = (σ3 σ1 ) ⊗ (σ3 σ1 ) = −σ2 ⊗ σ2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 92

92 Problems and Solutions

The neutral element is I2 ⊗ I2 . Each element is its own inverse. Thus the
set forms a group under matrix multiplication.

Problem 30. The spin matrices for spin-2 particles (for example graviton)
are given by

0 2 √0 0 0
 
2 √0 6 √0 0 
1
J1 =  0 6 √0 6 0,

2
0 0 6 0 2

0 0 0 2 0
0 −2 0 0 0
 
√
i
2 √0 − 6 0
√ 0 
J2 =  0 6 √0 − 6 0 ,
 
2
0 0 6 0 −2

0 0 0 2 0
2 0 0 0 0
 
0 1 0 0 0 
J3 =  0 0 0 0 0 .
 
0 0 0 −1 0
 
0 0 0 0 −2
(i) Show that the matrices are hermitian.
(ii) Find the eigenvalues and eigenvectors of these matrices.
(iii) Calculate the commutation relations.
(iv) Are the matrices unitary?

Solution 30. (i) Obviously the matrices are hermitian, i.e.

J1∗ = J1 , J2∗ = J2 , J3∗ = J3 .

(ii) The eigenvalues of J1 are −2, 2, −1, 1, 0 with the corresponding nor-
malized eigenvectors
1 1 1
     
−2  2  −1 
1√  1√  1
u1 =  6  , u2 =  6  , u3 =  0  ,
 
4 4 2
2 2 1
  
1 1 −1
1 1
   
1  √ 0
1 3 √ √ 

u4 =  0  , u5 = √  − 2/ 3  .
 
2 8
−1 0
 
−1 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 93

Matrix Properties 93

The eigenvectors form an orthonormal basis in C5 . The eigenvalues of J2

are 2, 1, 0, −1, −2 with the corresponding normalized eigenvectors

1 1
   
1

 −2i   2i
√  −i 
1 1 
 1− 6
 1
u1 =  √  , u2 =  , u3 =  0  ,
 
4 − 6 4  −2i  2
−i

2i 1
   
−1
1 1

1 1
   
i  √  0 
1 3 √ √ 
u4 =  0  , u5 = √  2/ 3  .
 
2 8 0 
i

−1 1
The eigenvalues of J3 are 2, 1, 0, −1, −2 with the corresponding eigenvec-
tors (standard basis)

1 0 0 0 0
         
0 1 0 0 0
u1 =  0  , u2 =  0  , u3 =  1  , u4 =  0  , u5 =  0  .
         
0 0 0 1 0
         
0 0 0 0 1

(iii) The commutation relations are [J1 , J2 ] = iJ3 , [J2 , J3 ] = iJ1 , [J3 , J1 ] =
iJ2 .
(iv) No the matrices are not unitary. Note that

det(J1 ) = det(J2 ) = det(J3 ) = 0

owing to the eigenvalue 0.

Problem 31. Two orthonormal bases in an n-dimensional complex Hilbert

space
{ |uj i : j = 1, 2, . . . , n }, { |vj i : j = 1, 2, . . . , n }
are called mutually unbiased if the inner products (scalar products) be-
tween all possible
√ pairs of vectors taken from distinct bases have the same
magnitude 1/ n, i.e.

1
|huj |vk i| = √ for all j, k ∈ { 1, 2, . . . , n }.
n

(i) Find such bases for the Hilbert space C2 .

(ii) Find such bases for the Hilbert space C3 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 94

94 Problems and Solutions

Solution 31. (i) As the first bases we select the standard basis
   
1 0
, .
0 1
For the second basis we could select
    
1 1 1 1
√ , √
2 1 2 −1
or     
1 1 1 1
√ , √ .
2 i 2 −i
Applying a unitary matrix to these two sets provide other such sets.
(ii) An example is the standard basis
     
 1 0 0 
0, 1, 0
0 0 1
 

and
      √ 
 1 1 −1
√ (−1 + i√3)/2 
1  1 
√ 1, √ (1 + 3)/2  , √ (−1 − i 3)/2  .
 3
1 3 (1 − √3)/2 3 1


Applying a unitary matrix to these two sets provide other such sets.

Problem 32. Find a 4 × 4 matrix A such that −A = A−1 = AT = A∗ .

Solution 32. Let 02 be the 2 × 2 zero matrix. We find

0 0 0 −1
 
 
02 −iσ2 0 0 1 0 
A= ≡ .
−iσ2 02 0 −1 0 0
1 0 0 0
This matrix plays a role for the charge conjugation in the Dirac equation.

Problem 33. Consider the spin matrices for a spin-1 particle

     
0 1 0 0 −i 0 1 0 0
1 1
S1 = √  1 0 1  , S2 = √  i 0 −i  , S3 =  0 0 0 
2 0 1 0 2 0 i 0 0 0 −1
and the unit vector n = ( sin(θ) cos(φ) sin(θ) sin(φ) cos(θ) ). We define
the scalar product
n · S := n1 S1 + n2 S2 + n3 S3 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 95

Matrix Properties 95

Consider the Hamilton operator

Ĥ = ~ω(n · S).

(i) Calculate the Hamilton operator Ĥ. Is the Hamilton operator Ĥ her-
mitian?
(ii) Calculate the trace of Ĥ.
(iii) Find the eigenvalues and normalized eigenvectors of Ĥ.
(iv) Do the eigenvectors form a basis in the Hilbert space C3 ?

Solution 33. (i) The Hamilton operator is given by

√
sin(θ)e−iφ / 2
 
cos(θ) √ 0 √
Ĥ = ~ω  sin(θ)eiφ / 2 0 √ sin(θ)e−iφ / 2  .
iφ
0 sin(θ)e / 2 − cos(θ)

Since exp(iφ) = exp(−iφ) the Hamilton operator is hermitian. Thus the

eigenvalues must be real.
(ii) The trace of Ĥ is 0. Thus the sum of the three eigenvalues of Ĥ must
be 0.
(iii) The eigenvalues of Ĥ are ~ω, 0, −~ω. The corresponding normalized
eigenvectors are
−iφ
√ 
− sin(θ)e−iφ / 2 −iφ
    
(1 + cos(θ))e
√ /2 (1 − cos(θ))e√ /2
 sin(θ)/ 2 ,  cos(θ) √  ,  − sin(θ)/ 2 .
(1 − cos(θ))eiφ /2 sin(θ)eiφ / 2 (1 + cos(θ))eiφ /2

(iv) The Hamilton operator is hermitian and the three eigenvalues are dif-
ferent. Thus the normalized eigenvectors form an orthonormal basis in the
Hilbert space C3 .

Problem 34. Consider a complex Hilbert space H and |φ1 i, |φ2 i ∈ H. Let
c1 , c2 ∈ C. An antilinear operator K in this Hilbert space H is characterized
by
K(c1 |φ1 i + c2 |φ2 i) = c∗1 K|φ1 i + c∗2 K|φ2 i.
A comb is an antilinear operator K with zero expectation value for all states
|ψi of a certain complex Hilbert space H. This means

hψ|K|ψi = hψ|LC|ψi = hψ|L|ψ ∗ i = 0

for all states |ψi ∈ H, where L is a linear operator and C is the complex
conjugation.
(i) Consider the two-dimensional Hilbert space H = C2 . Find a unitary
2 × 2 matrix such that hψ|U C|ψi = 0.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 96

96 Problems and Solutions

(ii) Consider the Pauli spin matrices with σ0 = I2 , σ1 , σ2 , σ3 . Find

3 X
X 3
hψ|σµ C|ψig µ,ν hψ|σν C|ψi
µ=0 ν=0

where g µ,ν = diag(−1, 1, 0, 1).

Solution 34. (i) We find U = σ2 since

ψ1∗
  
∗ 0 −i
hψ|σ2 C|ψi = hψ|σ2 |ψ i = ( ψ1∗ ψ2∗ ) = 0.
i 0 ψ2∗

(ii) We have
3 X
X 3
hψ|σµ C|ψig µ,ν hψ|σν C|ψi = −hψ|σ0 |ψ ∗ i2 + hψ|σ1 |ψ ∗ i2 + hψ|σ3 |ψ ∗ i2
µ=0 ν=0
= 0.

Problem 35. Consider the Hilbert space Cd . Let |ji (j = 1, . . . , d) be

an orthonormal basis in Cd . Then a d × d matrix A acting in Cd can be
written as
Xd
A= ajk |jihk|
j,k=1

with ajk ∈ C. Obviously A depends on the underlying orthonormal basis.

If we have the standard basis, then A reduces to the matrix A = (ajk ). We
2
can associate a vector |ψA i in the Hilbert space Cd with the matrix A via
d
X
|ψA i = ajk |ji ⊗ |ki.
j,k=1

(i) Let d = 2 and consider the standard basis

   
1 0
|1i = , |2i = .
0 1

Find A and |ψA i.

(ii) Let d = 2 and consider the Hadamard basis
   
1 1 1 1
|1i = √ , |2i = √ .
2 1 2 −1
Find A and |ψA i.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 97

Matrix Properties 97

Solution 35. (i) We have

A = a11 |1ih1| + a12 |1ih2| + a21 |2ih1| + a22 |2ih2|

         
a11 0 0 a12 0 0 0 0 a11 a12
= + + + =
0 0 0 0 a21 0 0 a22 a21 a22

and

|ψA i = a11 |1i ⊗ |1i + a12 |1i ⊗ |2i + a21 |2i ⊗ |1i + a22 |2i ⊗ |2i
a11 0 0 0 a11
         
 0   a12   0   0   a12 
= + + + = .
0 0 a21 0 a21
0 0 0 a22 a22
(ii) We have

A = a11 |1ih1| + a12 |1ih2| + a21 |2ih1| + a22 |2ih2|

       
1 1 1 1 1 −1 1 1 1 1 1 −1
= a11 + a12 + a21 + a22
2 1 1 2 1 −1 2 −1 −1 2 −1 1
 
1 a11 + a12 + a21 + a22 a11 − a12 + a21 − a22
=
2 a11 + a12 − a21 − a22 a11 − a12 − a21 + a22
and

|ψA i = a11 |1i ⊗ |1i + a12 |1i ⊗ |2i + a21 |2i ⊗ |1i + a22 |2i ⊗ |2i
1 1 1 1
       
1 1 1  −1  1  1  1  −1 
= a11   + a12   + a21   + a22 
2 1 2 1 2 −1 2 −1

1 −1 −1 1
a11 + a12 + a21 + a22
 
1  a − a12 + a21 − a22 
=  11 .
2 a11 + a12 − a21 − a22
a11 − a12 − a21 + a22
Extend to d = 3 and consider the orthonormal basis
     
1 0 1
1   1
|1i = √ 0 , |2i =  1  , |3i = √  0  .
2 1 0 2 −1

Find A and |ψA i. Describe the connection of the map A 7→ |ψA i with the
vec-operator.

Problem 36. Let s be a spin with a fixed total angular momentum

quantum number
s ∈ {1/2, 1, 3/2, 2, . . .}.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 98

98 Problems and Solutions

The (normalized) eigenstates of x3 -angular momentum |s, mi form a ladder

with
m = −s, −s + 1, . . . , s − 1, s.
The eigenstates |s, mi form an orthonormal basis in a 2s + 1 dimensional
Hilbert space. For example if s = 1/2 we have the two states |1/2, −1/2i,
|1/2, 1/2i and can identify
   
1 0
|1/2, 1/2i 7→ , |1/2, −1/2i 7→ .
0 1

Thus we have the Hilbert space C2 . For s = 1 we have the three states
|1, −1i, |1, 0i, |1, 1i and can identify
     
0 0 1
|1, −1i 7→  0  , |1, 0i 7→  1  , |1, 1i 7→  0  .
1 0 0
A spin coherent state |s, θ, φi for s = 1/2, 1, 3/2, . . . can be given by
s
m=s
X (2s)!
|s, θ, φi = (cos(θ/2))s+m (sin(θ/2))s−m e−imφ |s, mi.
m=−s
(s + m)!(s − m)!

(i) Find |1/2, θ, φi and write it as a vector in C2 .

(ii) Find |1, θ, φi and write it as a vector in C3 .
(iii) For a given s find the scalar product hs, m|s, θ, φi.

Solution 36. (i) We obtain

|1/2, θ, φi = sin(θ/2)eiφ/2 |1/2, −1/2i + cos(θ/2)e−iφ/2 |1/2, 1/2i.
Thus
cos(θ/2)e−iφ/2
 
|1/2, θ, φi 7→ .
sin(θ/2)eiφ/2
(ii) For s = 1 we obtain
√
sin2 (θ/2)eiφ |1, −1i + 2 cos(θ/2) sin(θ/2)|1, 0i + cos2 (θ/2)e−iφ |1, 1i.
Thus we the state in C3
−iφ
 2

√ cos (θ/2)e
|1, θ, φi 7→  2 cos(θ/2) sin(θ/2)  .
sin2 (θ/2)eiφ
(iii) Since hs, m0 |s, mi = δm0 ,m we obtain
s
(2s)!
hs, m|s, θ, φi = (cos(θ/2))s+m (sin(θ/2))s−m e−imφ .
(s + m)!(s − m)!
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 99

Matrix Properties 99

Problem 37. Let n ≥ 1 and {|0ii, |1i, . . . , |ni} be an orthonormal basis

in Cn+1 . Consider the linear operators ((n + 1) × (n + 1) matrices)
n
X n √
X
a†n =
p
an = j|j − 1ihj|, k|kihk − 1|.
j=1 k=1

Find the commutator [an , a†n ]. Note that

n
X
|`ih`| = In+1 .
`=0

Solution 37. We have

n
n X n
X p √ X
an a†n = j k|j − 1ihj|kihk − 1| = k|k − 1ihk − 1|
j=1 k=1 k=1
n X
X n √ n
X
a†n an =
p
k j|kihk − 1|j − 1ihj| = j|jihj|.
k=1 j=1 j=1

Thus [an , a†n ] = an a†n − a†n an = In+1 − (n + 1)|nihn|.

Problem 38. Let z ∈ C. Consider the spin-1 matrix

 
0 −i 0
1 
S2 = √ i 0 −i  .
2 0 i 0

Calculate exp(zS2 ). Then substitute z = −iωt.

Solution 38. Since S23 = S2 , S24 = S22 etc we obtain

z3 z5
 2
z4
  
z
exp(zS2 ) = I3 + S2 z + + + · · · + S22 + + ···
3! 5! 2! 4!
= I3 + S2 sinh(z) + S22 (cosh(z) − 1).

With z = −iωt we obtain

exp(−iωtS2 ) = I3 − i sin(ωt)S2 + (cos(ωt) − 1)S22 .

Since  
1 0 −1
1
S22 =  0 2 0 
2
−1 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 100

100 Problems and Solutions

we end up with
 √ 
1 + (cos(ωt) −√1)/2 − sin(ωt)/ 2 (− cos(ωt) +√1)/2
exp(−iωtS2 ) =  − sin(ωt)/ 2 cos(ωt)√ − sin(ωt)/ 2  .
(− cos(ωt) + 1)/2 sin(ωt)/ 2 1 + (cos(ωt) − 1)/2

Problem 39. Consider the Hilbert space M2 (C) of all 2 × 2 matrices over
C with scalar product

hA, Bi := tr(AB ∗ ), A, B ∈ M2 (C).

The standard basis is

       
1 0 0 1 0 0 0 0
E11 = , E12 = , E21 = , E22 = .
0 0 0 0 1 0 0 1

A mutually unbiased basis is

   
1 1 1 0 1 1 0 1
µ0 = √ σ0 = √ , µ1 = √ σ1 = √ ,
2 2 0 1 2 2 1 0
   
1 1 0 −i 1 1 1 0
µ2 = √ σ2 = √ , µ3 = √ σ3 = √ .
2 2 i 0 2 2 0 −1
(i) Express the Hadamard matrix
 
1 1 1
A= √
2 1 −1

with this mutually unbiased basis.

(ii) Express the Bell matrix

1 0 0 1
 
1 0 1 1 0 
B=√ 
2 0 1 −1 0

1 0 0 −1

with the basis (sixteen dimensional) given by µj ⊗ µk , (j, k = 0, 1, 2, 3).

Solution 39. (i) We have the expansion

3
X
A= hA, µj iµj
j=0

with
hA, µ0 i = 0, hA, µ1 i = 1, hA, µ2 i = 0, hA, µ3 i = 1.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 101

Matrix Properties 101

Hence A = µ1 + µ3 .
(ii) We have the expansion
3 X
X 3
B= hB, µj ⊗ µk i(µj ⊗ µk ).
j=0 k=0

The only nonzero expansion coefficients are

√ √
hB, µ3 ⊗ µ0 i = 2, hB, µ1 ⊗ µ1 i = 2.

Hence √ √
B= 2µ3 ⊗ µ0 + 2µ1 ⊗ µ1 .

Problem 40. Two orthonormal bases in an n-dimensional complex Hilbert

space
{ |uj i : j = 1, 2, . . . , n }, { |vj i : j = 1, 2, . . . , n }
are called mutually unbiased if inner products (scalar products) between all
possible
√ pairs of vectors taken from distinct bases have the same magnitude
1/ n, i.e.
1
|huj |vk i| = √ for all j, k ∈ { 1, 2, . . . , n }.
n

(i) Find such bases for the Hilbert space C2 . Start of with the standard
basis    
1 0
u1 = , u2 = .
0 1
(ii) Find such bases for the Hilbert space C3 . Start of with the standard
basis      
1 0 0
u1 =  0  , u2 =  1  , u3 =  0  .
0 0 1
(iii) Find such bases for the Hilbert space C4 using the result from C2 and
the Kronecker product.

Solution 40. (i) For the second basis we could select

   
1 1 1 1
v1 = √ , v2 = √ .
2 1 2 −1
Another selection would be
   
1 1 1 1
v1 = √ , v2 = √ .
2 i 2 −i
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 102

102 Problems and Solutions

(ii) A possible solution is

   √   √ 
1 1/ 3 √ 1/ 3 √
1
v1 = √  1  , v2 =  −i/2 − 1/(2√ 3)  , v3 =  i/2 − 1/(2 √3)  .
3 1 i/2 − 1/(2 3) −i/2 − 1/(2 3)

Problem 41. (i) Let A, B be n × n matrices over C such that A2 = In

and B 2 = In . Furthermore assume that

[A, B]+ ≡ AB + BA = 0n

i.e. the anticommutator vanishes. Let α, β ∈ C. Calculate eαA+βB using

∞
X (αA + βB)j
eαA+βB = .
j=0
j!

(ii) Consider the case that n = 2 and

 
1 0
α = −iωt, A = σ3 =
0 −1
 
0 1
β = −i∆t/~, B = σ1 = .
1 0
(iii) Consider the case that n = 8 and

α = −iωt, A = σ3 ⊗ σ3 ⊗ σ3

β = −i∆t/~, B = σ1 ⊗ σ1 ⊗ σ1 .

Solution 41. (i) Since BA = −AB we have

(αA + βB)2 = (α2 + β)2 In , (αA + βB)3 = (α2 + β 2 )(αA + βB).

Thus in general we have for positive n

(αA + βB)n = (α2 + β 2 )n/2 In for n even

and
(αA + βB)n = (α2 + β 2 )n/2−1 for n odd
Thus we have the expansion
1 2 1 1
eαA+βB = In (1 + (α + β 2 ) + (α2 + β 2 )2 + (α2 + β 2 )3 + · · ·)
2! 4! 6!
1 1
+(αA + βB)(1 + (α2 + β 2 ) + (α2 + β 2 )2 + · · ·).
3! 5!
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 103

Matrix Properties 103

This can be summed up to

p αA + βB p
eαA+βB = In cosh( α2 + β 2 ) + p sinh( α2 + β 2 ).
2
α +β 2

(ii) We have
p p it p 2 2
α2 + β 2 = −ω 2 t2 − ∆2 t2 /~2 = ~ ω + ∆2 .
~
√
We set E := ~2 ω 2 + ∆2 . It follows that
 
αA + βB ~ωA + ∆B −~ω/E −∆/E
=− = .
2
α +β 2 E −∆/E ~ω

Thus
 
αA + βB p −i sin(Et/~)~ω −i sin(Et/~)∆/E
sinh( α2 + β 2 ) =
−i sin(Et/~)∆/E i sin(Et/~)~ω
p
α2 + β 2

and
 
αA+βB cos(Et/~) − i sin(Et/~)~ω/E −i sin(Et/~)∆/E
e = .
−i sin(Et/~)∆/E cos(Et/~) + i sin(Et/~)~ω/E

Problem 42. Let H be an n × n hermitian matrix and λ1 , . . . , λn be the

eigenvalues with the pairwise orthogonal normalized eigenvectors v1 , . . . ,
vn . Then we can write
Xn
H= λ` v` v`∗ .
`=1

Let
P = In − vj vj∗ − vk vk∗ + vj vk∗ + vk vj∗ , j 6= k.
(i) What is the condition on the eigenvalues of H such that P HP ∗ = H.
(ii) Find P 2 .

Solution 42. (i) Note that P is hermitian. Utilizing v`∗ vj = δ`j we find
by straightforward calculation
n
X
P HP ∗ = λ` v` v`∗ + (λk − λj )vj vj∗ + (λj − λk )vk vk∗ .
`=1

Thus λj = λk .
(ii) We obtain P 2 = In .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 104

104 Problems and Solutions

Problem 43. Consider the orthogonal group O(n, R) ⊂ Rn×n with the
linear product (v, w ∈ O(n, R))
hv, wi := tr(v T w).
The orthogonal projection Π(g) : Rn×n → Tg O(n) is given by
1
Π(g)v := (v − gv T g).
2
(i) Let n = 2 and
   
0 1 a b
g= , v= .
1 0 c d
Find the orthogonal projection.
(ii) Let n = 2 and
   
0 1 a b
g= , v= .
1 0 c d
Find the orthogonal projection.

Solution 43. (i) We have

       
1 a b 0 1 a c 0 1 1 a−d 0
Π(g)v = − = .
2 c d 1 0 b d 1 0 2 0 d−a
(ii) We have
     
1 a b 0 1 a c 0 1
Π(g)v = −
2 c d −1 0 b d −1 0
 
1 a+d 0
= .
2 0 a+d
Note that   
a−d 0 a+d 0
tr = 0.
0 d−a 0 a+d
Consider the case that  
1 1 1
g=√ .
2 1 −1

Problem 44. Let n ≥ 1 and m ≥ 1. Consider the T = (tj1 ...jm order-m

tensor of size (n × · · · × n) (m-times), (j1 , . . . , jm = 1, . . . , n). One defines
the operator on v ∈ Cn written as
n
X n
X
(T vm−1 )k := ··· tkj2 ...jm vj2 . . . vjm , k = 1, . . . , n.
j2 =1 jm =1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 105

Matrix Properties 105

The (E−) eigenvector of T are the fixed points (up to scaling) of this
operator
T vm−1 = λv where v 6= 0.
Let m = 3, n = 2 with t122 = 1, t211 = 1 and all other entries are 0. Solve
the eigenvalue problem.

Solution 44. We obtain

(T v2 )1 = a122 v22 = v22 = λv1 , (T v2 )2 = a211 v12 = v12 = λv2

Now λ = 0 is not a solution, since v12 = v22 = 0 implies v1 = v2 = 0. If

λ 6= 0, then v1 6= 0 and v2 6= 0. We also have v13 = v23 .

Problem 45. Starting from Maxwell’s equations in vacuum

1 ∂E ∂B
curl(B) = , curl(E) = − , div(E) = 0, div(B) = 0
c2 ∂t ∂t
and Kramer’s vector F := E + icB, F∗ := E − icB show that the photon is
a spin-1 particle.

Solution 45. Using Kramer’s vector we can write

i ∂F i ∂F∗
curl(F) = , curl(F∗ ) = − , div(F) = 0, div(F∗ ) = 0.
c ∂t c ∂t
Let

 +1 if jk` are an even permutation of the integers 123
jk` = −1 if jk` are an odd permutation of the integers 123 .
0 otherwise


Since
3
3 X 3 3
X ∂ XX ∂
(curlF)j = jk` F` = − kj` F`
∂xk ∂xk
k=1 `=1 k=1 `=1
we can write
3 X
3  
X ∂ i ∂Fj
−i (−ikj` ) F` = .
∂xk c ∂t
k=1 `=1
Introducing the differential operator
∂
p̂k := −i
∂xk
we find
3 X
3
X i ∂F`
− p̂k ikj` F` = .
j=1 k=1
c ∂t
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 106

106 Problems and Solutions

For fixed k, −ikj` is a 3 × 3 matrix, Sk(j,`) . The equation for F then takes
the form !
3
X i ∂F
p̂k Sk F = (p̂ · S)F = .
c ∂t
k=1

Using the definition of jk` , we obtain the representation for the 3×3
matrices
     
0 0 0 0 0 1 0 −1 0
S1 = i  0 0 −1  , S2 = i  0 0 0  , S3 = i  1 0 0.
0 1 0 −1 0 0 0 0 0

The commutators are [S1 , S2 ] = iS3 , [S2 , S3 ] = iS1 , [S3 , S1 ] = iS2 . We

have S × S = iS and
S12 + S22 + S32 = 2I3
where I3 is the 3 × 3 identity matrix. Thus Maxwell’s equations describe a
particle of spin-1.

Programming Problems

Problem 1. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Find the eigen-

values and eigenvectors of the 8 × 8 hermitian matrix

H = (σ1 ⊗ σ1 + σ2 ⊗ σ2 + σ3 ⊗ σ3 ) ⊗ σ1 .

The 8 × 8 hermitian matrix can be written as a direct sum

0 −1 0 2
 
   
0 1  −1 0 2 0  0 1
H= ⊕ ⊕ .
1 0 0 2 0 −1 1 0

2 0 −1 0

Solution 1. Thus the eigenvalues can be calculated from the two 2 × 2

matrices and the 4 × 4 matrix. Applying the Maxima program
/* directsum.mac */
sig1: matrix([0,1],[1,0]);
sig2: matrix([0,-%i],[%i,0]);
sig3: matrix([1,0],[0,-1]);
sig11: kronecker_product(sig1,sig1);
sig22: kronecker_product(sig2,sig2);
sig33: kronecker_product(sig3,sig3);
S: sig11 + sig22 + sig33;
H: kronecker_product(S,sig1);
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 107

Matrix Properties 107

eigenvectors(H);
D: matrix([0,-1,0,2],[-1,0,2,0],[0,2,0,-1],[2,0,-1,0]);
eigenvectors(D);

we find the eigenvalues −1 (3 times), +1 (3 times), +3 (1 times), −3 (1

times) of the matrix H. The eigenvalues of the 4 × 4 matrix are −1, 1, −3,
3 and the eigenvectors are

1 1 1 1
       
 −1  1  1   −1 
,  , , .
1 1 −1 −1
  
−1 1 −1 1

Problem 2. Let In be the n × n identity matrix. An invertible matrix

2 2
X ∈ Cn ×n satisfies the Yang-Baxter equation if

(X ⊗ In )(In ⊗ X)(X ⊗ In ) = (In ⊗ X)(X ⊗ In )(In ⊗ X).

If X satisfies the Yang-Baxter equation, then X ∗ satisfies the Yang-Baxter

equation. If X satisfies the Yang-Baxter equation, then X −1 satisfies the
Yang-Baxter equation. If X satisfies the Yang-Baxter equation and Q ∈
Cn×n is an arbitrary invertible matrix. Then

e = (Q ⊗ Q)X(Q ⊗ Q)−1
X

also satisfies the Yang-Baxter equation. Show that

1 0 0 1
 
(1 + i)  0 1 1 0
X=
2 0 −1 1 0
 
−1 0 0 1

satisfies the Yang-Baxter equation with n = 2.

Solution 2. The following Maxima program provides the proof

/* YBBell.mac */
I2: matrix([1,0],[0,1]);
X: ((1+%i)/2)*matrix([1,0,0,1],[0,1,1,0],[0,-1,1,0],[-1,0,0,1]);
T1: kronecker_product(X,I2);
T2: kronecker_product(I2,X);
F: (T1 . T2 . T1) - (T2 . T1 . T2);
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 108

108 Problems and Solutions

3.3 Supplementary Problems

Problem 1. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Consider the

sixteen 4 × 4 matrices
   
σ1 0 2 σ2 02
I16 , Γ1 = , Γ2 = ,
02 σ1 02 σ2
     
02 σ3 02 −iσ3 −σ3 02
Γ3 = , Γ4 = , Γ5 = ,
σ3 02 iσ3 02 02 σ3
1
Γ[µ,ν] = i(Γµ Γν − Γν Γµ ), µ, ν = 1, 2, 3, 4, 5.
2
Show that
Γν Γµ + Γµ Γν = 2δνµ I4
[Γν , Γ[λ,µ] ] = 2iδνλ Γµ − 2iδνµ Γλ
[Γ[ν,µ] , Γ[λ,σ] ] = 2iδµλ Γ[ν,σ] − 2iδνλ Γ[µ,σ] + 2iδνσ Γ[ν,λ] − 2iδµσ Γ[ν,λ] .
Do the 16 matrices form an orthonormal basis in the Hilbert space of the
4 × 4 matrices?

Problem 2. Consider the 25 × 25 matrices

A = I2 ⊗ I2 ⊗ σ1 ⊗ σ3 ⊗ σ3
B = I2 ⊗ I2 ⊗ σ2 ⊗ σ3 ⊗ σ3
C = σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 .

Find A2 , B 2 , C 2 , [A, B], [B, C], [C, A], [A, B]+ , [B, C]+ , [C, A]+ .

Problem 3. Let A be an n × n matrix over C and f : C → C be analytic

in a region D containing the spectrum of A. Then the matrix f (A) can be
defined as the Cauchy integral formula
Z
1
f (A) = (zIn − A)−1 f (z)dz.
2πi ∂D

Let  
0 0 1
A = 0 1 0
1 0 0
with the spectrum +1 (twice) and −1. Find exp(A) applying the Cauchy
integral formula.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 109

Matrix Properties 109

Problem 4. Let α > 0 and A be an n × n matrix over C. Show that

Z
−αA i
e = e−αλ (A − λIn )−1 dλ
2π C

where C is the contour in the complex λ plane which encloses all eigenvalues
of the matrix A. Let  
0 1
A= .
0 0
Calculate the right-hand side.

Problem 5. Show that the vector is normalized in R4

cos(θ1 )
 
sin(θ1 ) cos(θ2 )
.
 
sin(θ1 ) sin(θ2 ) cos(θ3 )

sin(θ1 ) sin(θ2 ) sin(θ3 )

Problem 6. Show that the equation of a hyperplane passing through the

points x1 , x2 , . . . , xn in Rn can be given in the form
 
1 1 1 ··· 1
det = 0.
x x1 x2 · · · xn

Apply it to n = 4 with (Bell basis)

1 1 0 0
       
1 0 1  0  1 1 1  1 
x1 = √   , x2 = √   , x3 = √   , x4 = √  .
2 0 2 0 2 1 2 −1
1 −1 0 0

Problem 7. Let σ1 , σ2 , σ3 be the Pauli spin matrices.

(i) Show that the matrices

1 1
(I2 + σj ), (I2 − σj ), j = 1, 2, 3
2 2
are projection matrices.
(ii) Show that the matrices

1 1
(I4 + σj ⊗ σk ), (I4 − σj ⊗ σk ), j, k = 1, 2, 3
2 2
are projection matrices.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 110

110 Problems and Solutions

(iii) Show that the matrices

1 1
(I8 + σj ⊗ σk ⊗ σ` ), (I8 − σj ⊗ σk ⊗ σ` ), j, k, ` = 1, 2, 3
2 2
are projection matrices.
(iv) Let U be an n × n unitary and hermitian matrix. Show that
1
Π= (In − U )
2
is a projection matrix. Show that Π ⊗ Π is a projection matrix.

Problem 8. Consider the matrices

   
0 1 0 1 0 1
A = 1 0 1, B = 0 1 0.
0 1 0 1 0 1

Write down all six 3 × 3 permutation matrices with

   
1 0 0 0 0 1
P0 =  0 1 0  , P5 =  0 1 0  .
0 0 1 1 0 0

Find the permutation matrices in this set such that Pj APjT = A. Find the
permutation matrices in this set such that Pj BPjT = B.

Problem 9. Consider the 4 × 4 matrix

−1/2 1/2 1/2 1/2
 
 1/2 −1/2 1/2 1/2 
A= .
1/2 1/2 −1/2 1/2
1/2 1/2 1/2 −1/2
Find the eigenvalues of A without calculating the eigenvalues. Utilize the
information from A2 , tr(A) and that the matrix A is symmetric over R.
Then find the eigenvalues of A ⊗ A and A ⊗ I4 + I4 ⊗ A.

Problem 10. Find the eigenvalues and normalized eigenvectors of the

Hamilton operator (16 × 16 hermitian matrix)

Ĥ = ~ω1 (σ3 ⊗ σ3 ⊗ I2 ⊗ I2 + I2 ⊗ I2 ⊗ σ3 ⊗ σ3 ) + ~ω2 (σ1 ⊗ σ1 ⊗ σ1 ⊗ σ1 ).

Problem 11. Consider the Hamilton operator

Ĥ = ~ω1 (σ3 ⊗ I2 + I2 ⊗ σ3 ) + ~ω2 σ1 ⊗ σ1 + ~ω3 σ2 ⊗ σ2 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 111

Matrix Properties 111

Find the eigenvalues and normalized eigenvectors of Ĥ.

Problem 12. Let n ≥ 2 and ω = e2πi/n . Consider the n × n matrices

0 0 ... 1
 
1 0 ... 0
D = diag(1 ω · · · ω n−1 ), Γ = 
 ... . . . ... 0  .


0 ... 1 0

So Γ is a permutation matrix with Γn = In . Furthermore Dn = In . Find

the commutator [D, Γ].

Problem 13. Let v0 , v1 , v2 , v3 be an orthonormal basis in the Hilbert

space C4 . Show that the vectors
1 1
u0 = (v0 + v1 + v2 + v3 ), u1 = (v0 − v1 + v2 − v3 ),
2 2
1 1
u2 = (v0 + v1 − v2 − v3 ), u3 = (v0 − v1 − v2 + v3 )
2 2
also form an orthonormal basis in C4 .

Problem 14. Let x ∈ R. Are the vectors

1 0 0 0
       
x  1   0  0
v1 =  2  , v2 =   , v3 =   , v4 =  
x 2x 2 0
x3 3x2 6x 6

linearly independent? Find Gram’s matrix

G = (vjT vk ), j, k = 1, 2, 3, 4

and its determinant.

Problem 15. Let  

cos(θ)
v(θ) = .
sin(θ)
Find all 2 × 2 matrices A, B such that v∗ ABv = (v∗ Av)(v∗ Bv).

Problem 16. Consider the symmetric binary matrices

   
0 1 0 1 0 1
A = 1 0 1, B = 0 1 0.
0 1 0 1 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 112

112 Problems and Solutions

Find the eigenvalues and eigenvectors of A and B. Find the eigenvalues

and eigenvectors of the anti-commutator [A, B]+ . Discuss.

Problem 17. Find all n × n matrices A and B such that

eA ⊗ eB = eA⊗B .

Problem 18. Consider the 2 × 2 elementary matrices

       
1 0 0 0 0 1 0 0
E11 = , E22 = , E12 = , E21 =
0 0 0 1 0 0 1 0

with the commutator [E12 , E21 ] = E11 − E22 . Let θ ∈ R. Show that

exp (−θ(E12 + E21 )) =

exp(− tanh(θ)E12 ) exp(ln(cosh(θ))(E22 − E11 )) exp(− tanh(θ)E21 ).

Problem 19. Let µ ∈ C, S the spin (S = 0, 1/2, 1, 3/2, 2, . . .) and |0i,

|1i, . . . , |2Si be the standard basis in C2S+1 . The Bloch coherent states |µi
are defined by
2S  1/2
1 X (2S)!
|µi = µp |pi.
(1 + |µ|2 )S p=0 p!(2S − p)!

Show that
2S  1/2
1 X (2S)!
hµ| = (µ∗ )p hp|.
(1 + |µ|2 )S p=0 p!(2S − p)!

Show that (completeness relation)

d2 µ
Z
1 + 2S
2 2
|µihµ| = I2S+1
π C (1 + |µ| )

where d2 µ = d(<(µ))(d(=(µ)). Show that the scalar product of two Bloch

coherent states is given as

(1 + ν ∗ µ)2S
hν|µi = .
(1 + |ν|2 )S (1 + |µ|2 )S

Let S = 1/2. Show that

1
|µi = (|0i + µ|1i).
(1 + |µ|2 )1/2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 113

Matrix Properties 113

Problem 20. (i) Let A, B be n × n matrices over C and C = A + B.

Show that (Trotter formula)
 n
e−τ C = lim e−τ A/n e−τ B/n , τ ≥ 0.
n→∞

(ii) Let A, B be n × n matrices. Show that using exponential theory such

that

exp(τ (A + B)) = exp(τ A/2) exp(τ B) exp(τ A/2) + O(τ 3 ).

Problem 21. Let s = 1/2, 1, 3/2, 2, . . . be the spin and let m = −s, −s +
1, . . . , s − 1, s. The simple Lie algebra su(2) has generators {S3 , S+ , S− }
with commutation relations

[S+ , S− ] = 2S3 , [S3 , S± ] = ±S± .

Given a finite-dimensional module with highest weight s the action of S3 ,

S− , S+ on the weight basis is

S3 |s, mi = m|s, mi eigenvalue equation

p
S+ |s, mi = (s − m)(s + m + 1)|s, m + 1i
p
S− |s, mi = (s + m)(s − m + 1)|s, m − 1i.

Study the eigenvalue problem of the Hamilton operators

H1 = ~ω(S1 ⊗ S1 + S2 ⊗ S2 + S3 ⊗ S3 )

H2 = ~ω(S1 ⊗ S2 + S2 ⊗ S3 + S3 ⊗ S1 )
with the basis

{ |s, m1 i ⊗ |s, m2 i : mj = −s, −s + 1, . . . , s − 1, s; j = 1, 2 }.

Problem 22. Consider C8 ∼

= C2 ⊗ C2 ⊗ C2 and
1 T
|ψi = √ ( 1 −1 1 −1 1 −1 1 −1 )
8
Find the Schmidt decomposition.

Problem 23. Consider the four Bell states

1 0
   
1 0 1 1
|ψ1 i = √   , |ψ2 i = √   ,
2 0 2 1
1 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 114

114 Problems and Solutions

0 1
  
1  1  1  0 
|ψ3 i = √  , |ψ4 i = √ 
2 −1 0

2
0 −1
and the Pauli spin matrices σ1 , σ2 , σ3 . Find

hψj |σk ⊗ σ` |ψj i

for j = 1, 2, 3, 4 and k, ` = 1, 2, 3. Apply Computer Algebra.

Problem 24. Consider the normalized (column) vector v ∈ Cn and the

n × n matrix M = In − 2vv∗ , where In is the n × n identity matrix. Show
that the matrix is hermitian and unitary.

Problem 25. Let c†1 , c†2 , c1 , c2 be Fermi creation and annihilation oper-
ators, respectively. Let α ∈ R. Show that

U (α) = exp(iα(c†1 c2 + c†2 c1 ))

= I + i sin(α)(c†1 c2 + c†2 c1 ) + (cos(α) − 1)(N̂1 + N̂2 − 2N̂1 N̂2 )

where N̂1 = c†1 c1 , N̂2 = c†2 c2 . First show that

(c†1 c2 + c†2 c1 )2 = N̂1 + N̂2 − 2N̂1 N̂2

(c†1 c2 + c†2 c1 )3 = c†1 c2 + c†2 c1
(c†1 c2 + c†2 c1 )4 = N̂1 + N̂2 − 2N̂1 N̂2 .

Show that N̂1 + N̂2 − 2N̂1 N̂2 is a projection operator.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 115

Chapter 4

Density Operators

4.1 Introduction
A density operator ρ or density matrix is a positive semidefinite operator on
a Hilbert space with unit trace. An operator is positive semidefinite if it is
hermitian and none of its (necessarily real) eigenvalues are less than zero.
The state of a quantum-mechanical system is characterized by a density
operator ρ with tr(ρ) = 1. The expectation value of an observable  (self-
adjoint operator, hermitian matrix), determined in an experiment as the
average value hÂi is given by

hÂi := tr(Âρ).

A density matrix or density operator is used in quantum theory to describe

the statistical state of a quantum system. If we have a pure state |ψi in a
Hilbert space then
ρ = |ψihψ|
defines a density matrix with (hψ|ψi = 1)

ρ2 = |ψihψ|ψihψ| = |ψihψ| = ρ.

For a mixed state we have the spectral representation

n
X
ρ= pj |ψj ihψj |
j=1

115
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 116

116 Problems and Solutions

where pj ≥ 0 for j = 1, 2, . . . , n,

n
X
pj = 1 and hψj |ψk i = δjk .
j=1

Then for the expectation value of a hermitian operator  in the Hilbert

space we have
Xn
tr(ρÂ) = pj hψj |Â|ψj i.
j=1

Let M be an arbitrary nonzero n × n matrix over C. Then

MM∗
ρ=
tr(M M ∗ )

is a density matrix.
If ρ1 and ρ2 are density matrices, then ρ1 ⊗ ρ2 , ρ1 ⊕ ρ2 are density matrices.
If ρ1 , ρ2 are pure states, then ρ1 ⊗ ρ2 and ρ1 ⊕ ρ2 are pure states.
If ρ is an n×n density matrix and U is an n×n unitary matrix (U −1 = U ∗ ),
then U ρU ∗ is a density matrix.
The eigenvalues of an n × n density matrix which is a pure state are 1 and
0 (n − 1 times), since ρ2 = ρ, tr(ρ) = 1 and ρ∗ = ρ.
Consider a quantum system of spin-1/2 particles. The density matrix de-
scribing the spin degree of freedom is a 2 × 2 matrix which can be written
as
1 1
ρ(n) = (I2 + n · σ) ≡ (I2 + n1 σ1 + n2 σ2 + n3 σ3 )
2 2
where σ1 , σ2 , σ3 denote the Pauli spin matrices and |n| ≤ 1. For |n| = 1
the density matrix describes a pure state, whereas for |n| < 1 one has a
mixed state. The density matrix ρ is thus uniquely determined by a point
of the unit sphere |n| ≤ 1.
The variance of an observable  and a density operator ρ in a Hilbert space
H is defined as
V (ρ, Â) := tr(ρÂ2 ) − (tr(ρÂ))2 .

Let |ψi be a normalized state in the Hilbert space H. If ρ = |ψihψ| (pure

state) we obtain

V (|ψihψ|, Â) = hψ|Â2 |ψi − hψ|Â|ψi2 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 117

Density Operators 117

4.2 Solved Problems

Problem 1. Is the 2 × 2 matrix
 
1/2 −1/2
ρ=
−1/2 1/2

a density matrix? If so is it a pure state? If so find the normalized state in

C2 that provides this density matrix.

Solution 1. We find that tr(ρ) = 1 and the matrix is hermitian over C.

The eigenvalues of ρ are 0 and 1. Thus ρ is a density matrix and a pure
state. The state  
1 1
|ψi = √
2 −1
is a normalized state that provides the density matrix ρ = |ψihψ|.

Problem 2. Let
 
cos(θ)
|ψi = , θ, φ ∈ R.
eiφ sin(θ)

be a normalized state in C2 . Does ρ := |ψihψ| define a density matrix?

Solution 2. We find the 2 × 2 matrix for ρ

e−iφ cos(θ) sin(θ)

 
cos2 (θ)
ρ = |ψihψ| = .
eiφ cos(θ) sin(θ) sin2 (θ)

Now ρ = ρ∗ , tr(ρ) = cos2 (θ) + sin2 (θ) = 1 and ρ2 = ρ. Hence we have a

density matrix (pure state).

Problem 3. Let r ≥ 0. Is the 2 × 2 matrix

1 1 + r cos(θ) r sin(θ)e−iφ
 
ρ=
2 r sin(θ)eiφ 1 − r cos(θ)
a density matrix? What are the conditions on r, θ, φ?

Solution 3. We have that tr(ρ) = 1 and the matrix is hermitian. Thus

the eigenvalues are real. The eigenvalues are given by
1 1
λ1,2 = ± r.
2 2
Thus the condition that ρ is density matrix is r ≤ 1. There is no condition
on φ and θ.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 118

118 Problems and Solutions

Problem 4. Let σ1 , σ2 and σ3 be the Pauli spin matrices. Let σ =

(σ1 , σ2 , σ3 ) and r ∈ R3 with r2 ≤ 1. Consider the 2 × 2 matrix (density
matrix)
1
ρ := (I2 + r · σ)
2
where r · σ := r1 σ1 + r2 σ2 + r3 σ3 . Let n be an arbitrary unit length vector
in R3 , i.e. n2 ≡ n21 + n22 + n23 = 1. Calculate tr((n · σ)ρ) i.e. we calculate
the expectation value of n · σ. Give an interpretation of the result.

Solution 4. We have
 
1 1
tr((n · σ)ρ) = tr (n · σ) (I2 + r · σ) = tr(n · σ + (n · σ)(r · σ))
2 2
 
3
1 1 X
= tr((n · σ)(r · σ)) = tr ni rj σi σj 
2 2 i,j=1
3
! 3
1 X X
= tr ni ri I2 = ni ri = n · r.
2 i=1 i=1

The vector r can be thought of as an expectation value of spin polarization,

and it can be obtained by measuring n · σ along each direction e1 , e2 and
e3 .

Problem 5. Let A be a nonzero n × n matrix over C. Then tr(AA∗ ) > 0.

Consider the map
AA∗
A→ρ= .
tr(AA∗ )
(i) Show that ρ is a density matrix.
(ii) Show that ρ is invariant under the map A → AU , where U is an n × n
unitary matrix.
(iii) Is AA∗ = A∗ A in general? A matrix is called a normal matrix if
AA∗ = A∗ A.
(iv) Consider the map
A∗ A
A→σ= .
tr(A∗ A)
Is σ = ρ?

Solution 5. (i) Since A is a nonzero matrix we find that AA∗ is nonzero

and tr(AA∗ ) 6= 0. The matrix ρ is positive-semidefinite and tr(ρ) = 1.
(ii) We set A0 = AU . Thus we have
0
0A0 A ∗ (AU )(AU )∗ AU U ∗ A∗ AA∗
ρ→ρ = 0∗ = = =
tr(A0 A ) tr((AU )(AU )∗ ) tr(AU U ∗ A∗ ) tr(AA∗ )
=ρ
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 119

Density Operators 119

where we used that U U ∗ = I.

(iii) In general we have AA∗ 6= A∗ A. For example, let
   
0 1 ∗ 0 0
A= , A = .
0 0 1 0

Then    
1 0 0 0
AA∗ = , A∗ A = .
0 0 0 1
Thus AA∗ 6= A∗ A. However, we have tr(AA∗ ) = tr(A∗ A).
(iv) From (iii) it follows that in general we have ρ 6= σ.

Problem 6. Find a normalized state |φi in the Hilbert space C2 such

that we have the density matrix
 
1 1
|φihφ| = I2 + √ (σ1 + σ3 ) .
2 2

Solution 6. We have to solve

 √ √ 
1 1 + 1/
√ 2 1/ 2√
|φihφ| = .
2 1/ 2 1 − 1/ 2

We obtain the normalized state

q √ √ 
( 2 + 1)/(2 2)
|φi =  q √ √ .
( 2 − 1)/(2 2)

Problem 7. Consider the 2 × 2 matrix

 √ −iφ 
√ 3/4 2e /4
ρ= .
2eiφ /4 1/4

(i) Is the matrix a density matrix?

(ii) If so do we have a pure state or a mixed state?
(iii) Find the eigenvalues of ρ.
(iv) Find tr(σ1 ρ), where σ1 is the first Pauli spin matrix.

Solution 7. (i) We have tr(ρ) = 1 and the matrix is hermitian. Further-

more the eigenvalues are nonnegative. Thus we have a density matrix.
(ii) Since ρ2 6= ρ we have a mixed state.
√ √
(iii) The eigenvalues are λ1 = (2 + 3)/4, λ2 = (2 − 3)/4 which also
indicate the state is mixed.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 120

120 Problems and Solutions

(iv) We obtain
1
tr(σ1 ρ) = √ cos(φ).
2

Problem 8. (i) The Hilbert-Schmidt distance between any two density

operators ρ1 and ρ2 is given by the Frobenius-Hilbert-Schmidt norm of their
differences p
DHS (ρ1 , ρ2 ) := tr((ρ1 − ρ2 )2 ).
Let    
1 0 0 0
ρ1 = , ρ2 = .
0 0 0 1
Calculate DHS (ρ1 , ρ2 ).
(ii) The Bures distance in the space of mixed quantum states described by
the density matrices ρ1 and ρ2 is defined as
q
1/2 1/2
DB (ρ1 , ρ2 ) := 2(1 − tr((ρ1 ρ2 ρ1 )1/2 )).

Let    
1/2 0 3/4 0
ρ1 = , ρ2 = .
0 1/2 0 1/4
Calculate the Bures distance DB (ρ1 , ρ2 ).
(iii) Let ρ, σ be two density operators acting in the same finite dimensional
Hilbert space. The trace distance between ρ and σ is defined as
1 p ∗
D(ρ, σ) := tr( (ρ − σ ∗ )(ρ − σ)).
2
Let
1/2 0 0 1/2 1/2 0 0 0
   
 0 0 0 0   0 0 0 0 
ρ= , σ= .
0 0 0 0 0 0 0 0
1/2 0 0 1/2 0 0 0 1/2
Find the trace distance.

Solution 8. (i) Since

 
−1/4 0
ρ1 − ρ2 =
0 1/4
√
we find DHS (ρ1 , ρ2 ) = 1/ 8.
(ii) Since  
1/2 1/2 3/8 0
ρ1 ρ2 ρ1 =
0 1/8
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 121

Density Operators 121

we find s √ √
2 2− 3−1
DB (ρ1 , ρ2 ) = √ .
2
(iii) We have
0 0 0 1/2
 
 0 0 0 0 
ρ−σ = .
0 0 0 0
1/2 0 0 0
Thus
1/4 0 0 0
 
0 0 0 0 
(ρ∗ − σ ∗ )(ρ − σ) =  .

0 0 0 0
0 0 0 1/4
It follows that D(ρ, σ) = 1/2.

Problem 9. Consider the linear operator (4 × 4 matrix) in the Hilbert

space C4
1
(1 − )I4 + (|0i ⊗ |0i)(h0| ⊗ h0|)
ρ=
4
where  is a real parameter with  ∈ [0, 1] and the state
 
1
|0i = .
0
Does ρ define a density matrix?

Solution 9. We find the diagonal matrix for ρ

(1 − )/4 +  0 0 0
 
0 (1 − )/4 0 0
ρ= .
 
0 0 (1 − )/4 0
0 0 0 (1 − )/4
Thus ρ = ρ∗ , tr(ρ) = 1, and hx|ρ|xi ≥ 0, for all x ∈ C4 . The last property
follows since all entries on the diagonal are non-negative. Thus ρ defines a
density matrix.

Problem 10. A mixed state is a statistical mixture of pure states, i.e.

the state is described by pairs of probabilities and pure states. Given a
mixture { (p1 , |ψ1 i), . . . , (pn , |ψn i) } we define its density matrix to be the
positive hermitian matrix
n
X
ρ= pj |ψj ihψj |
j=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 122

122 Problems and Solutions

where the pure states |ψj i are normalized (i.e. hψj |ψj i = 1), and pj ≥ 0
for j = 1, 2, . . . , n with p1 + p2 + · · · + pn = 1.
(i) Find the probability that measurement in the orthonormal basis

{ |k1 i, . . . , |kn i}

will yield |kj i.

(ii) Find the density matrix ρU when the mixture is transformed according
to the unitary matrix U .

Solution 10. (i) From the probability distribution of states in the mix-
ture we have for the probability P (kj ) of measuring the state |kj i (j =
1, 2, . . . , n)
n
X n
X
P (kj ) = pl |hkj |ψl i|2 = pl hkj |ψl ihψl |kj i = hkj |ρ|kj i.
l=1 l=1

(ii) After applying the transform U to the states in the mixture we have
the new mixture { (p1 , U |ψ1 i), . . . , (pn , U |ψn i) }, with the density matrix
 
Xn Xn
∗
ρU = pj U |ψj ihψj |U = U  pj |ψj ihψj | U ∗ = U ρU ∗ .
j=1 j=1

Problem 11. (i) The Bell state

1
|ψi = √ (|0i ⊗ |0i + |1i ⊗ |1i)
2
has the density matrix

1 0 0 1
 
1 0 0 0 0
ρ=  .
2 0 0 0 0
1 0 0 1

Show that ρ can be written as linear combination of the matrices Λ00 =

1 1 1 1
2 (I2 ⊗ I2 ), Λ11 = 2 (σ1 ⊗ σ1 ), Λ22 = 2 (σ2 ⊗ σ2 ) and Λ33 = 2 (σ3 ⊗ σ3 ).
(ii) The Werner state is described by the density matrix

(1 − x)/4 0 0 0
 
0 (1 + x)/4 −x/2 0
ρW =
 
0 −x/2 (1 + x)/4 0

0 0 0 (1 − x)/4
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 123

Density Operators 123

where x ∈ [0, 1]. Show that ρW can also be written as linear combination
of the operators given in (i).

Solution 11. (i) We find the linear combination

1 1 1 1
ρ= Λ00 + Λ11 − Λ22 + Λ33 .
2 2 2 2
(ii) We obtain the linear combination
1 x x x
ρW = Λ00 − Λ11 − Λ22 − Λ33 .
2 2 2 2

Problem 12. Suppose we expand a density matrix for N qubits in terms

of Kronecker products of Pauli spin matrices
3 3 3
1 X X X
ρ= . . . cj0 j1 ...jN −1 σj0 ⊗ σj1 ⊗ · · · ⊗ σjN −1
2N j =0 j =0 j =0
0 1 N −1

where σ0 = I2 .
(i) What is condition on the expansion coefficients if we impose ρ∗ = ρ?
(ii) What is the condition on the expansion coefficients if we impose tr(ρ) =
1?
(iii) Calculate tr(ρσk0 ⊗ σk1 ⊗ · · · ⊗ σkN −1 ).

Solution 12. (i) Since σ1 = σ1∗ , σ2 = σ2∗ , σ3 = σ3∗ and I2 = I2∗ we find
that the expansion coefficients are real.
(ii) Since tr(A ⊗ B) = tr(A)tr(B) for square matrices A and B and
tr(σ1 ) = tr(σ2 ) = tr(σ3 ) = 0, tr(I2 ) = 2
we find c00...0 = 1.
(iii) Since tr(σ1 σ2 ) = 0, tr(σ2 σ3 ) = 0, tr(σ3 σ1 ) = 0 we find
tr(ρσk0 ⊗ σk1 ⊗ · · · ⊗ σkN −1 ) = ck0 k1 ...kN −1 .

Problem 13. Let A and B be a pair of qubits and let the density matrix
of the pair be ρAB , which may be pure or mixed. We define the spin flipped
density matrix to be
ρeAB := (σ2 ⊗ σ2 )ρ∗AB (σ2 ⊗ σ2 )
where the asterisk denotes complex conjugation and transpose in the stan-
dard basis
{ |0i ⊗ |0i, |0i ⊗ |1i, |1i ⊗ |0i, |1i ⊗ |1i }
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 124

124 Problems and Solutions

and  
0 −i
σ2 = .
i 0
Since both ρAB and ρeAB are positive operators, it follows that the prod-
uct ρAB ρeAB , though non-hermitian, also has only real and non-negative
eigenvalues. Consider the Bell state
1
|ψi := √ (|0i ⊗ |0i + |1i ⊗ |1i)
2
and ρ := |ψihψ|. Find the eigenvalues of ρAB ρeAB .

Solution 13. Since

1 0 0 1
 
1 0 0 0 0
ρ = |ψihψ| = 
2 0 0 0 0

1 0 0 1
we have ρ∗ = ρ. Furthermore
0 0 0 −1
 
 0 0 1 0 
σ2 ⊗ σ2 =  .
0 1 0 0
−1 0 0 0
Thus ρe = ρ and ρe
ρ = ρ with eigenvalues 1, 0, 0, 0. The tangle of the density
matrix ρAB is defined as
τAB := [max { µ1 − µ2 − µ3 − µ4 , 0}]2
where µj are the square root of the eigenvalues of ρAB ρeAB ordered in de-
creasing order. For the special case in which the state of AB is pure, the
matrix ρAB ρeAB has only one non-zero eigenvalue. One can show that
τAB = 4 det(ρA )
where ρA is the density matrix of qubit A, that is, the trace of ρAB over
qubit B.

Problem 14. Consider the density matrix

1
ρ= (I2 + r · σ)
2
where r · σ := r1 σ1 + r2 σ2 + r3 σ3 and r2 ≤ 1. Consider the four normalized
vectors a1 , a2 , a3 , a4 in R3
       
1 1 −1 −1
1   1  1 1
a1 = √ 1 , a2 = √ −1  , a3 = √  1  , a4 = √  −1 
3 1 3 −1 3 −1 3 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 125

Density Operators 125

such that 
4 1 1 for j = k
aTj ak = δjk − = .
3 3 −1/3 for j 6= k
We have
4 4
X 3X
aj = 0, aj aTj = I3 .
j=1
4 j=1
Such a quartet of vectors consists of the vectors pointing from the center
of a cube to nonadjacent corners. These four vectors can be viewed as
the normal vectors for the faces of the tetrahedron that is defined by the
other four corners of the cube. Owing to the conditions the four vectors are
normalized. Each such quartets of aj ’s defines a positive operator-valued
measure for minimal four-state tomography owing to
4
X 1
Pj = I 2 , Pj := (I2 + aj · σ) .
j=1
4

(i) Show that pj := hPj i = tr(Pj ρ) = 41 (1 + aj · r).

(ii) Given pj for j = 1, 2, 3, 4 find the density matrix ρ.

Solution 14. (i) Since tr(σ1 ) = tr(σ2 ) = tr(σ3 ) = 0 we have

1
hPj i = tr(Pj ρ) = tr((I2 + aj · σ)(I2 + r · σ))
8
1 1
= tr(I2 + aj · σ + r · σ + (aj · σ)(r · σ)) = tr(I2 + (aj · σ)(r · σ))
8 8
1
= tr(I2 + (aj · r)I2 )
8
1
= (1 + aj · r).
4
(ii) Since  
4
3 X
I3 r =  aj aTj  r
4 j=1

the vector r is obtained as

4 4 4
3X X 1 X
r= (aj · r)aj = 3 (1 + aj · r) aj = 3 pj aj
4 j=1 j=1
4 j=1
P4
where we used j=1 aj = 0. From aj · σ = 4Pj − I2 and substituting r
from above yields the density matrix
4 4
X X 1
ρ=6 pj Pj − I2 = hPj i(6Pj − I2 ) = (I2 + r · σ).
j=1 j=1
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 126

126 Problems and Solutions

It follows that pj is restricted to the range 0 ≤ pj ≤ 1/2 and the probabili-

ties pj obey the inequalities
4
1 X 2 3 + r2 1
≤ pj = ≤ .
4 j=1 12 3

The upper bound is reached by all pure states, ρ = ρ2 and r2 = 1. The

lower bound is reached for the completely mixed state, ρ = 13 I2 and r2 = 0.

Problem 15. Let |0i, |1i be the standard basis in C2 . Consider the
mixed states    
1 3 4 1 3 4
|0i + |1i + |0i − |1i
2 5 5 2 5 5
and
9 16
{ |0i } + { |1i }.
25 25
Find the density matrices. Discuss.

Solution 15. In the first case we have

     
1 9/25 12/25 1 9/25 −12/25 9/25 0
ρ1 = + = .
2 12/25 16/25 2 −12/25 16/25 0 16/25
In the second case we have
     
9 1 0 16 0 0 9/25 0
ρ2 = + = .
25 0 0 25 0 1 0 16/25
Thus these two different mixed states correspond to the same density matrix
and thus they are indistinguishable.

Problem 16. Let ρ1 and ρ2 be n × n density matrices. Let λj denote

the eigenvalues of ρ1 − ρ2 with corresponding orthonormal eigenvectors |φj i
where j = 1, 2, . . . , n.
(i) Find the difference |D1 − D2 | between the probability distributions
D1 and D2 for the measurement of the mixtures ρ1 and ρ2 in the basis
{ |φ1 i, . . . , |φn i }.
(ii) Show that measurement in the basis { |φ1 i, . . . , |φn i } maximizes the
difference |D1 − D2 |.
Hint. Use Schur’s theorem. For any hermitian matrix A, let

a1 ≥ a2 ≥ · · · ≥ an

be the non increasing diagonal entries of A and

µ1 ≥ µ2 ≥ · · · ≥ µn
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 127

Density Operators 127

the non increasing eigenvalues of A. Then for 1 ≤ k ≤ n

k
X k
X
µj ≥ aj
j=1 j=1

where equality holds for k = n.

Solution 16. (i) We write ρ1 and ρ2 in the basis { |φ1 i, . . . , |φn i }. In

this basis we have
n
X n
X n
X
|D1 −D2 | = |hφj |ρ1 |φj i−hφj |ρ2 |φj i| = |hφj |(ρ1 −ρ2 )|φj i| = |λj |.
j=1 j=1 j=1

(ii) Let U be an arbitrary unitary transform (change of basis). We define

P := U ρ1 U ∗ and Q := U ρ2 U ∗ . The matrix P − Q is hermitian. Let
q1 ≥ q2 ≥ · · · ≥ qn
denote the non decreasing diagonal entries of P − Q in the { |φ1 i, . . . , |φn i }
basis and
ν1 ≥ ν2 ≥ · · · ≥ νn
be the non decreasing eigenvalues (i.e. λj ) of P − Q. Consider the differ-
ence |D10 − D20 | between the probability distributions D10 and D20 for the
measurement of the mixtures ρ1 and ρ2 in the basis { U |φ1 i, . . . , U |φn i }
n
X
|D10 − D20 | = |hφj |U ∗ ρ1 U |φj i − hφj |U ∗ ρ2 U |φj i|
j=1
Xn n
X
= |hφj |P |φj i − hφj |Q|φj i| = |qj |.
j=1 j=1

Since tr(P − Q) = tr(P ) − tr(Q) = 1 − 1 = 0 and

n
X n
X
tr(P ) − tr(Q) = (hφj |P |φj i − hφj |Q|φj i) = qj
j=1 j=1

we have for all 1 ≤ k ≤ n

   
 k   n 
X   X 

 q j
 = 
 qj .

 j=1  j=k+1 

We conclude from the triangle inequality that

 
n
 k
X X 

|qj | ≥ 2 
 qj 
j=1  j=1 
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 128

128 Problems and Solutions

where equality holds for some 1 ≤ k0 ≤ n. Similarly

 
Xn  k 
X 
|νj | ≥ 2  νj  .
j=1  j=1 

From Schur’s theorem we have

n
X k0
X k0
X n
X
|νj | ≥ νj ≥ qj = |qj |.
j=1 j=1 j=1 j=1

Thus
n
X n
X
|νj | = |D1 − D2 | ≥ |D10 − D20 | = |qj |.
j=1 j=1

Problem 17. Let σ1 , σ2 , σ3 be the Pauli spin matrices.

(i) Is the 4 × 4 matrix ρ = 14 (I2 ⊗ I2 + σ1 ⊗ σ1 ) a density matrix?
(ii) Is the 4 × 4 matrix ρ = 14 (I4 − σ1 ⊗ σ1 − σ2 ⊗ σ2 − σ3 ⊗ σ3 ) a density
matrix?

Solution 17. (i) We have tr(ρ) = 1 and ρ is hermitian. The eigenvalues

of ρ are 1/2, 1/2, 0, 0. Thus the matrix ρ is a density matrix (mixed state).
(ii) Obviously we have ρ∗ = ρ, tr(ρ) = 1. Furthermore the matrix is positive
semidefinite. We have ρ2 = ρ. Thus ρ is a density matrix. We have a pure
state. The density matrix is given by
0 0 0 0
 
 0 1/2 −1/2 0 
ρ= .
0 −1/2 1/2 0
0 0 0 0

Problem 18. Consider the eight Gell-Mann matrices

     
0 1 0 0 −i 0 1 0 0
G1 =  1 0 0  , G2 =  i 0 0  , G3 =  0 −1 0  ,
0 0 0 0 0 0 0 0 0
     
0 0 1 0 0 −i 0 0 0
G4 =  0 0 0  , G5 =  0 0 0  , G6 =  0 0 1  ,
1 0 0 i 0 0 0 1 0
   
0 0 0 1 0 0
1
G7 =  0 0 −i  , G8 = √  0 1 0  .
0 i 0 3 0 0 −2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 129

Density Operators 129

They all have trace 0 and they are hermitian.

(i) Find the anticommutation relations for these matrices.
(ii) Consider the matrix
 
8
1 √ X
ρ= I3 + 3 nj Gj 
3 j=1

where nj ∈ R. What is the condition on the vector n = (n1 n2 . . . n8 )T

such that ρ is a density matrix for a pure state?

Solution 18. (i) We obtain for the anticommutators

8
4 X
[Gj , Gk ]+ = δjk I3 + 2 djk` G`
3
`=1

where the nonzero components of the completely symmetric tensor djk` are

1 1
d118 = d228 = d338 = −d888 = √ , d448 = d558 = d668 = d778 = − √
3 2 3
1
d146 = d157 = −d247 = d256 = d344 = d355 = −d366 = −d377 = .
2
(ii) The conditions for a density matrix ρ of a pure state are

ρ∗ = ρ, ρ2 = ρ, tr(ρ) = 1.

Imposing these conditions we obtain for the vector n that

n = n, nT n = 1, n?n=n

where
8
√ X
(a ? b)j := 3 djk` ak b` .
k,`=1

Problem 19. Let ρ denote the density matrix (mixed state)

 
1 1 0
ρ :=
2 0 1

in C2 . Find a pure state |Ψi ∈ C2 ⊗ C2 such that the reduced density

matrix found by taking the partial trace over the second system (C2 ) is ρ.
In other words purify the density matrix ρ to obtain a pure state |Ψi.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 130

130 Problems and Solutions

Solution 19. We begin with the Schmidt decomposition of |Ψi over the
Hilbert space C2 ⊗ C2
Sch(|Ψi,C2 ,C2 )
X p
|Ψi = λj |ψj i ⊗ |φj i
j=1

where λ1 and λ2 are the eigenvalues of ρ and |ψ1 i and |ψ2 i are the cor-
responding orthonormal eigenvectors of ρ. The states |φ1 i and |φ2 i in C2
are also orthonormal. The eigenvalues and eigenvectors of ρ are given by
λ1 = λ2 = 1/2 and
   
1 0
|ψ1 i = , |ψ2 i = .
0 1
Thus the spectral decomposition of ρ is given by
   
1 1 1 0
ρ= (1 0) + (0 1).
2 0 2 1
Hence    
1 1 1 0
|Ψi = √ ⊗ |φ1 i + √ ⊗ |φ2 i
2 0 2 1
where hφ1 |φ1 i = hφ2 |φ2 i = 1 and hφ1 |φ2 i = hφ2 |φ1 i = 0. Thus we could
take |Ψi as one of the Bell states
1 1 0 0
       
1 0 1  0  1 1 1  1 
√  , √  , √  , √ 
2 0 0 2 1 2 −1

2
1 −1 0 0
but not a product state.

Problem 20. Let r ∈ [0, 1]. Consider the density matrix

ρ = r|Φ+ ihΦ+ | + (1 − r)|00ih00|
where |Φ+ i is the Bell state
1
|Φ+ i = √ (|00i + |11i).
2
Calculate the eigenvalues of ρ.

Solution 20. The matrix representation of ρ is

1 − r/2 0 0 r/2
 
0 0 0 0 
ρ= .

0 0 0 0
r/2 0 0 r/2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 131

Density Operators 131

The characteristic equation is

 r  r  r r
1− − λ (−λ)(−λ) − λ − (−λ)(−λ) = 0.
2 2 2 2
Thus two eigenvalues are 0 (λ2 = 0, λ3 = 0) with the corresponding eigen-
vectors
0 0
   
1 0
u2 =   , u3 =   .
0 1
0 0
The characteristic equation reduces to
 r r  r2
1− −λ −λ − =0
2 2 4
with the eigenvalues
1 1p
λ1,4 = ± 1 + 2r(r − 1).
2 2
If r = 0 the eigenvalues reduce to 1 and 0. If r = 1 the eigenvalues also
reduce to 1 and 0.

(1)
Problem 21. Let ρj (j = 1, 2, . . . , n) be density matrices in a finite-
(2)
dimensional Hilbert space H1 . Let ρj (j = 1, 2, . . . , n) be density matrices
in a finite-dimensional Hilbert space H2 . Show that the convex combination
n n
(1) (2)
X X
ρ= λj ρj ⊗ ρj , λj ≥ 0, λj = 1
j=1 j=1

is also a density matrix.

Solution 21. We have

 
n n
(1) (2) (1) (2)
X X
tr(ρ) = tr  λ j ρj ⊗ ρj  = λj tr(ρj ⊗ ρj )
j=1 j=1
n n
(1) (2)
X X
= λj tr(ρj )tr(ρj ) = λj
j=1 j=1
= 1.
(1) (2)
Obviously ρ ≥ 0, since ρj ≥ 0, ρj ≥ 0 and λj ≥ 0 for j = 1, 2, . . . , n.

Problem 22. Given the Schrödinger equation

∂
i~ |ψi = Ĥ|ψi.
∂t
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 132

132 Problems and Solutions

Find the time-evolution of the density matrix

n
X
ρ(t) := |ψ (j) (t)ihψ (j) (t)|.
j=1

Solution 22. From the Schrödinger equation we find

∂ (j)
−i~ hψ (t)| = hψ (j) (t)|Ĥ.
∂t
Thus
n    
∂ρ X ∂ (j) ∂ (j)
= |ψ (t)i hψ (j) (t)| + |ψ (j) (t)i hψ (t)|
∂t j=1 ∂t ∂t
n
1 X    
= Ĥ|ψ (j) (t)i hψ (j) (t)| − |ψ (j) (t)i hψ (j) (t)|Ĥ
i~ j=1
1
= (Ĥρ(t) − ρ(t)Ĥ)
i~
1
= [Ĥ, ρ(t)].
i~
Note that the equation of motion for ρ(t) differs from the Heisenberg equa-
tion of motion by a minus sign. Since ρ(t) is constructed from state vectors
it is not an observable like other hermitian operators, so there is no reason
to expect that its time-evolution will be the same. The solution to the
equation of motion is given by

ρ(t) = e−iĤt/~ ρ(0)eiĤt/~ .

Problem 23. Consider the Hamilton operator

γ γ
Ĥ(t) = − σ · B(t) ≡ − (σ1 B1 (t) + σ2 B2 (t) + σ3 B3 (t))
2 2
where γ denotes the gyromagnetic ratio and B(t) denotes the time-dependent
magnetic induction. The time-evolution of the density matrix ρ(t) obeys
the von Neumann equation
dρ(t)
i~ = [Ĥ(t), ρ(t)]
dt
and the time-dependent expectation value of the spin vector is given by

hσ(t)i := tr(σρ(t))
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 133

Density Operators 133

or, written in components

hσ1 (t)i = tr(σ1 ρ(t)), hσ2 (t)i = tr(σ2 ρ(t)), hσ3 (t)i = tr(σ3 ρ(t)).

It follows that the Bloch vector n(t) pertaining to ρ(t) is related to the spin
vector as follows
n(t) = hσ(t)i
or, written in components

n1 (t) = hσ1 (t)i, n2 (t) = hσ2 (t)i, n3 (t) = hσ3 (t)i.

Find the time-evolution of n(t).

Solution 23. We have

   
dnj dσj (t) dρ(t)
= = tr σj
dt dt dt

where j = 1, 2, 3. Inserting the right-hand side of the von Neumann equa-

tion, using the cyclic invariance of the trace

tr(XY Z) = tr(ZXY ) = tr(Y ZX)

and the properties σ1 σ2 = iσ3 , σ2 σ3 = iσ1 , σ3 σ1 = iσ2 , we obtain

d γ
n(t) = n(t) × B(t)
dt ~
where × denotes the vector product.

Problem 24. Consider the state

 
cos(θ)
|ψi =
eiφ sin(θ)

in the Hilbert space C2 , where φ, θ ∈ R. Let ρ(t = 0) = ρ(0) = |ψihψ| be

a density matrix at time t = 0. Given the Hamilton operator Ĥ = ~ωσ1 .
Solve the von Neumann equation to find ρ(t).

Solution 24. We obtain for the density matrix at time t = 0

e−iφ cos(θ) sin(θ)

 
cos2 (θ)
ρ(0) = |ψihψ| = .
eiφ cos(θ) sin(θ) sin2 (θ)

Now
ρ(t) = e−iĤt/~ ρ(0)eiĤt/~ = e−iωtσ1 ρ(0)eiωtσ1 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 134

134 Problems and Solutions

Since  
cos(ωt) i sin(ωt)
eiĤt/~ = eiωtσ1 =
i sin(ωt) cos(ωt)
it follows that
e−iφ cos(θ) sin(θ)
 
∗ cos2 (θ)
ρ(t) = U (t) U (t)
iφ
e cos(θ) sin(θ) sin2 (θ)

where  
cos(ωt) i sin(ωt)
U (t) = .
i sin(ωt) cos(ωt)

Problem 25. Consider the Hilbert space Cn . Let ρ be a density matrix,

i.e. ρ ≥ 0 and tr(ρ) = 1. The mean value of an observable A (hermitian
n × n matrix) is given by
hAi = tr(ρA).
If the density ρ is unknown, then it may be determined using n2 mean values
hA(k) i (k = 1, 2, . . . , n2 ) obtained from measurement if the set {A(k) } is a
basis in the space of all hermitian n × n matrices.
(i) Let n = 2,  
0 −i
A = σ2 =
i 0
and tr(ρA) = 0, tr(ρA2 ) = 1, tr(ρA3 ) = 0, tr(ρA4 ) = 1. Find the density
matrix.
(ii) Let n = 2 and tr(ρI2 ) = 1, tr(ρσ1 ) = −1, tr(ρσ2 ) = 0, tr(ρσ3 ) = 0.
Find ρ.

Solution 25. (i) Note that σ22 = I2 , σ23 = σ2 , σ24 = I2 . The density
matrix is  
1− 0
ρ= ,  ∈ [0, 1].
0 
(ii) The 2×2 matrices I2 , σ1 , σ2 , σ3 form an orthogonal basis in the Hilbert
space of the 2 × 2 matrices with scalar product hX, Y i = tr(XY ∗ ). The
density matrix is  
1/2 −1/2
ρ= .
−1/2 1/2

Problem 26. Let |0i, |1i be the standard basis in C2 . Consider the
entangled state
1
|ψi = √ (|0i ⊗ |1i − |1i ⊗ |0i)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 135

Density Operators 135

with the density matrix ρ = |ψihψ|. Find the reduced density matrix ρ1 .
Discuss.

Solution 26. We obtain

 
1 1 1 0
ρ1 = (|0ih0| + |1ih1|) = .
2 2 0 1

Thus we have mixed state.

Problem 27. Consider a mixture of 25% of the pure state (1, 0)T , 25%
of the pure state (0, 1)T and 50% of the pure state √12 (1, 1)T described by
the density matrix
     
1 1 1 0 1 1 1 1
ρ= (1 0) + (0 1) + √ √ (1 1).
4 0 4 1 2 2 1 2

Find the spectral representation of ρ. Use the spectral representation of ρ to

find another mixture of pure states with the same (measurement) statistical
properties as ρ.

Solution 27. The density matrix takes the form

 
1 2 1
ρ=
4 1 2

with the eigenvalues λ1 = 3/4 and λ2 = 1/4 with the corresponding nor-
malized eigenvectors
   
1 1 1 1
v1 = √ , v2 = √ .
2 1 2 −1

Applying the spectral theorem ρ can be written as

   
3 1 1 1 1 1 1 1
ρ= √ √ (1 1) + √ √ ( 1 −1 ) .
4 2 1 2 4 2 −1 2
√
by a mixture of 75% of the state (1, 1)T / 2
Consequently ρ can be realized √
and 25% of the state (1, −1)T / 2.

Problem 28. Consider the Hilbert space Cn . Let ρ be a density matrix

in this Hilbert space and H and K be two hermitian n × n matrices. One
defines hHi := tr(ρH), hH 2 i := tr(ρH 2 ) and analogously for K. Let
p p
∆H := hH 2 i − hHi2 , ∆K := hK 2 i − hKi2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 136

136 Problems and Solutions

Then we have the uncertainty relation

1
(∆H)(∆K) ≥ |hi[H, K]i| .
2
Let
     
1 0 0 0 1 0 0 i 0
1
ρ = 0 0 0, H = 1 2 0, K =  −i 0 0  .
2
0 0 1 0 0 0 0 0 0

Show that the uncertainty relation becomes an equality for the given ρ, H
and K.

Solution 28. Note that ρ is a mixed state since ρ2 6= ρ. Straightforward

calculations yield
1 1
hHi = 0, hKi = 0, hH 2 i = , hK 2 i =
2 2
and the commutator of H and K is given by
 
−2i 2i 0
[H, K] =  2i 2i 0  .
0 0 0

Thus |hi[H, K]i| = 1/2 and the equality follows.

Programming Problems

Problem 1. Consider the matrix

√
2e−iφ /4
 
3/4
ρ = √ iφ .
2e /4 1/4

Check that the matrix is a density matrix. Is it a pure or mixed state?

Apply computer algebra. Find tr(σ1 ρ).

Solution 1. The following Maxima program will do the job

/* densitycheck.mac */
load("nchrpl");
rho: matrix([3/4,sqrt(2)*exp(-%i*phi)/4],[sqrt(2)*exp(%i*phi)/4,1/4]);
rhoT: transpose(rho); rhoTC: conjugate(rhoT);
tr: mattrace(rho);
E: eigenvalues(rho);
e1: first(first(E)); e2: second(first(E));
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 137

Density Operators 137

if(rho=rhoTC and tr=1 and e1>0 and e2>0)

then print("matrix is density matrix")
else print("matrix is not a density matrix");
rho2: rho . rho;
if(rho2=rho and tr=1 and rho=rhoTC) then print("pure state")
else print("not a pure state");
sig1: matrix([0,1],[1,0]);
trsig1: mattrace(rho . sig1);

The matrix
√ is a density
√ matrix and a mixed state. The two eigenvalues are
(2 + 3)/4, (2 − 3)/4. We find tr(σ1 ρ) = √12 cos(φ).

Problem 2. Consider the density matrix

 
1 1 −1
ρ=
2 −1 1

and let A be an 2 × 2 real symmetric matrix. Assume that tr(ρA) = −1,

tr(ρA2 ) = 1. Reconstruct the matrix from this information.

Solution 2. Utilizing the Maxima program

/* density1.mac */
load("nchrpl");
rho: matrix([1/2,-1/2],[-1/2,1/2]);
A: matrix([a11,a12],[a12,a22]);
A2: A . A;
r1: mattrace(rho . A); r1: ratsimp(r1);
r2: mattrace(rho . A2); r2: ratsimp(r2);
solve([r1+1=0,r2-1=0],[a11,a12,a22]);

we obtain  
r 1+r
A=
1+r r
with r an arbitrary real constant.

Problem 3. Let S be the set of unit vectors in the Hilbert space Cn .

Let u ∈ S. A function µ(u) from S to R is called a generalized probability
measure if the following two conditions hold: (i) for u ∈ S, 0 ≤ µ(u) ≤ 1,
n
Pn if u1 , . . . , un form an orthonormal basis in the Hilbert space C , then
(ii)
j=1 µ(uj ) = 1.
Let n ≥ 3. Then any generalized probability measure µ on Cn has the form

µ(ρ) = tr(ρuu∗ )

for a uniquely defined density matrix ρ (Gleason 1957).

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 138

138 Problems and Solutions

(i) Consider the Hilbert space C3 , the orthonormal basis

     
1 0 1
1   1
u1 = √ 0 , u2 =  1  , u3 = √  0 
2 1 0 2 −1

and the density matrix

 
1 1 1
1
ρ= 1 1 1.
3
1 1 1

Find µ(u1 ), µ(u2 ), µ(u3 ).

(ii) Consider the Hilbert space C4 , the orthonormal basis

eiφ eiφ
  
1  0  1  0 
u1 = √  , u2 = √  ,
2 0 2 0
eiφ −eiφ

0 0
   
1  eiφ  1  eiφ 
u3 = √  iφ  , u4 = √  iφ 
2 e 2 −e
0 0
and the density matrix

1 1 1 1
 
1 1 1 1 1
ρ=  .
4 1 1 1 1
1 1 1 1

Find µ(u1 ), µ(u2 ), µ(u3 ), µ(u4 ).

Solution 3. Applying the Maxima program

/* Gleason.mac */
load("nchrpl");
u1: matrix([1],[0],[1])/sqrt(2); u1T: transpose(u1);
u2: matrix([0],[1],[0]); u2T: transpose(u2);
u3: matrix([1],[0],[-1])/sqrt(2); u3T: transpose(u3);
rho: matrix([1,1,1],[1,1,1],[1,1,1])/3;
muu1: mattrace(rho . u1 . u1T);
muu2: mattrace(rho . u2 . u2T);
muu3: mattrace(rho . u3 . u3T);
v1: matrix([exp(%i*phi)],[0],[0],[exp(%i*phi)])/sqrt(2);
v1T: transpose(v1); v1TC: conjugate(v1T);
v2: matrix([exp(%i*phi)],[0],[0],[exp(%i*phi)])/sqrt(2);
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 139

Density Operators 139

v2T: transpose(v2); v2TC: conjugate(v2T);

v3: matrix([0],[exp(%i*phi)],[exp(%i*phi)],[0])/sqrt(2);
v3T: transpose(v3); v3TC: conjugate(v3T);
v4: matrix([0],[exp(%i*phi)],[exp(%i*phi)],[0])/sqrt(2);
v4T: transpose(v4); v4TC: conjugate(v4T);
rho4: matrix([1,1,1,1],[1,1,1,1],[1,1,1,1],[1,1,1,1])/4;
muv1: mattrace(rho4 . v1 . v1TC); muv2: mattrace(rho4 . v2 . v2TC);
muv3: mattrace(rho4 . v3 . v3TC); muv4: mattrace(rho4 . v4 . v4TC);

we find for (i) µ(u1 ) = 2/3, µ(u2 ) = 1/3, µ(u3 ) = 0 and for (ii) we find
µ(u1 ) = 1/2, µ(u2 ) = 1/2, µ(u3 ) = 1/2, µ(u4 ) = 1/2.

4.3 Supplementary Problems

Problem 1. Consider a spin-1 system. Any pure state can be parametrized,
with a suitable choice of its phase as
 
eiα sin(θ) cos(φ)
|ψi =  eiβ sin(θ) sin(φ) 
cos(θ)

where 0 ≤ θ, φ ≤ π/2 and 0 ≤ α, β < 2π. Find the density matrix ρ =

|ψihψ| and tr(ρS1 ), where S1 is the spin-1 matrix
 
1 0 1
1 
S1 = √ 0 1 0.
2 1 0 1

Note that the density matrix depends on all four parameters.

Problem 2. Let σ0 , σ1 , σ2 , σ3 be the Pauli spin matrices, where σ0 = I2

is the 2 × 2 unit matrix and let
T
v = ( v1 v2 v3 )

be a vector in R3 with kvk ≤ 1.

(i) Show that ρv = 21 (σ0 + v1 σ1 + v2 σ2 + v3 σ3 ) is a density matrix.
P3
(ii) Is ρ = 14 (σ0 ⊗ σ0 + j=1 vj σj ⊗ σj ) a density matrix?
(iii) Is
3
1 X
ρ= (σ0 ⊗ σ 0 ⊗ σ 0 + vj σ j ⊗ σ j ⊗ σ j )
23 j=1

a density matrix? Extend the result to n Kronecker products.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 140

140 Problems and Solutions

Problem 3. Let |0i, |1i, . . . , |ni be an orthonormal basis in Cn+1 . Are

the states
n
1 1 X
|ψ0 i = √ |0i ⊗ |0i + √ |ji ⊗ |ji
2 2n j=1
n
1 1 X
|ψ1 i = √ |0i ⊗ |0i − √ |ji ⊗ |ji
2 2n j=1

normalized? Are the state orthogonal to each other? Is

ρ = (|ψ0 ihψ0 |) ⊗ (|ψ1 ihψ1 |)

a density matrix?

Problem 4. Consider the Pauli spin matrices σ1 and σ2 . Let ρ be a 2 × 2

density matrix with tr(ρσ1 ) = 1, tr(ρσ2 ) = 0. Reconstruct the density
matrix ρ from this information. Show that
 
1 1 1
ρ= .
2 1 1

Problem 5. Let H1 and H2 be two Hilbert spaces and H1 ⊗ H2 be the

product Hilbert space. Let ρ be a density operators of the Hilbert space
H1 ⊗ H2 . Show that if one of the reduced density operators trH2 (ρ) = ρ1
or trH1 (ρ) = ρ2 is pure, then ρ = ρ1 ⊗ ρ2 . If both ρ1 and ρ2 are pure, then
ρ is pure too.

Problem 6. Find a normalized state |ψi in the Hilbert space C2 such

that we have the density matrix
 
1 1
|ψihψ| = I2 + √ (σ1 + σ2 + σ3 ) .
2 3
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 141

Chapter 5

Trace and Partial Trace

5.1 Introduction
Let H be the finite dimensional Hilbert space Cn with an orthonormal basis
{ |φj i : j = 1, 2, . . . , n }. Let A be a linear operator (n × n matrix) acting
in this Hilbert space. Then the trace of A is defined as
n
X
tr(A) := hφj |A|φj i .
j=1

The trace is independent of the chosen orthonormal basis. For the trace we
have cyclic invariance. Let A, B, C be n × n matrices over C. Then
tr(AB) = tr(BA)
and (cyclic invariance)
tr(ABC) = tr(CAB) = tr(BCA).
The trace of an n × n matrix A is the sum of the eigenvalues counting
multiplicities. The eigenvalues of A can be reconstructed from
tr(A) = λ1 + λ2 + · · · + λn
tr(A2 ) = λ21 + λ22 + · · · + λ2n
..
.
tr(An ) = λn1 + λn2 + · · · + λnn .

141
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 142

142 Problems and Solutions

If |ψi is a normalized state in Cn , then

tr(|ψihψ|) = 1.
For any n × n matrix over C we have the identity
det(exp(A)) = exp(tr(A)).
Let A be an n × n matrix over C and B be an m × m matrix over C. Then
tr(A ⊗ B) = tr(A)tr(B).
The calculation of the partial trace plays a central role in quantum com-
puting. Suppose a finite dimensional quantum system SAB is a system
composed of two subsystems SA and SB . The finite dimensional Hilbert
space H of SAB is given by the tensor product of the individual Hilbert
spaces HA ⊗ HB . Let NA := dim(HA ) and NB := dim(HB ). Let ρAB
be the density matrix of SAB . Using the partial trace we can define the
density operators ρA and ρB in the subspaces HA and HB as follows
NB
X
ρA := trB (ρAB ) ≡ (IA ⊗ hφj |)ρAB (IA ⊗ |φj i)
j=1

and
NA
X
ρB := trA (ρAB ) ≡ (hψj | ⊗ IB )ρAB (|ψj i ⊗ IB i)
j=1
where IA is the identity operator in HA , IB is the identity operator in HB
and
|φj i, (j = 1, 2, . . . , NB )
is an orthonormal basis in HB and
|ψj i, (j = 1, 2, . . . , NA )
is an orthonormal basis in HA . For example we could select the standard
bases in the two finite dimensional Hilbert spaces HA and HB .

The partial trace can also be calculated as follows. Consider a bipartite

state
n−1
X n−1
X n−1
X n−1
X n−1
X n−1
X
|ψi = cjk |jki ≡ cjk |ji ⊗ |ki , cjk c∗jk = 1
j=0 k=0 j=0 k=0 j=0 k=0

in the finite-dimensional Hilbert space H = C ⊗ Cn . We can define the

n

n × n matrix
Λjk := cjk , j, k = 0, 1, . . . , n − 1.
Then we have (prove it)
ρA = trB (ρ) = trB (|ψihψ|) = ΛΛ† .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 143

Trace and Partial Trace 143

5.2 Solved Problems

Problem 1. Consider the Pauli spin matrix σ1 . Calculate the trace of σ1
with respect to the standard basis. Calculate the trace of σ1 with respect
to the Hadamard basis
    
1 1 1 1
√ , √ .
2 1 2 −1

Solution 1. Obviously for the standard basis we find tr(σ1 ) = 0. Since

       
1 0 1 1 1 1 0 1 1 1
√ (1 1) √ = 1, √ (1 − 1) √ = −1
2 1 0 2 1 2 1 0 2 −1
we also obtain (as expected since the trace is independent of the chosen
orthonormal basis) for the Hadamard basis the result 0. In the Hadamard
basis we have the Pauli spin matrix σ3 .

Problem 2. Consider the hermitian matrix

 
1 1 1
H=√ .
2 1 −1
Find tr(H) and tr(H 2 ) and then the eigenvalues from this information.

Solution 2. Since tr(H) = 0, tr(H 2 ) = 2 and λ1 + λ2 = 0, λ21 + λ22 = 2

we obtain the eigenvalues λ1 = 1, λ2 = −1.

Problem 3. Consider the entangled state in C4

0
 
       
1 1 0 0 1 1  1 
|ψi = √ ⊗ − ⊗ =√ 
0 1 1 0 2 −1

2
0
and the density matrix ρ = |ψihψ|. Find ρ1 = tr2 (ρ), ρ2 = tr1 (ρ) i.e.
calculate the partial trace.

Solution 3. We have
1
hψ| = √ ( 0 1 −1 0).
2
Thus
0 0 0 0
 
1 0 1 −1 0
ρ=  .
2 0 −1 1 0
0 0 0 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 144

144 Problems and Solutions

Using the basis

       
1 1 0 0 1 0
⊗ , ⊗
0 0 1 1 0 1
and the basis
       
1 0 1 1 0 0
⊗ , ⊗
0 1 0 0 1 1
we find the density matrix (mixed state)
 
1 1 0
ρ1 = ρ2 = .
2 0 1

Problem 4. Consider the 4 × 4 matrix (density matrix)

|uihu| = (uj ūk ), j, k = 1, . . . , 4
in the product Hilbert space HA ⊗ HB ≡ C4 , where HA = HB = C2 .
(i) Calculate trA (|uihu|), where the basis is given by
   
1 0
⊗ I2 , ⊗ I2
0 1
and I2 denotes the 2 × 2 unit matrix.
(ii) Find the partial trace trB (|uihu|), where the basis is given by
   
1 0
I2 ⊗ , I2 ⊗ .
0 1

Solution 4. (i) Since

1 0 0 0
   
   
1 0 1 0 0 0
⊗ I2 =  , ⊗ I2 = 
0 0 0 1 1 0

0 0 0 1
we find, using the transpose of these matrices on the left-hand side of |uihu|,
that
 u1 ū1 u1 ū2 u1 ū3 u1 ū4 1 0
  

1 0 0 0  u2 ū1 u2 ū2 u2 ū3 u2 ū4   0 1 
trA (|uihu|) =
0 1 0 0 u3 ū1 u3 ū2 u3 ū3 u3 ū4 0 0
  
u4 ū1 u4 ū2 u4 ū3 u4 ū4 0 0
 u1 ū1 u1 ū2 u1 ū3 u1 ū4 0 0
  

0 0 1 0  u2 ū1 u2 ū2 u2 ū3 u2 ū4   0 0 
+ .
0 0 0 1 u3 ū1 u3 ū2 u3 ū3 u3 ū4 1 0
 
u4 ū1 u4 ū2 u4 ū3 u4 ū4 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 145

Trace and Partial Trace 145

Using matrix multiplication and matrix addition we obtain

 
u1 ū1 + u3 ū3 u1 ū2 + u3 ū4
trA (|uihu|) = .
u2 ū1 + u4 ū3 u2 ū2 + u4 ū4

(ii) Since

1 0 0 0
   
   
1 0 0 0 1 0
I2 ⊗ = , I2 ⊗ =
0 0 1 1 0 0

0 0 0 1

we find

u1 ū1 u1 ū2 u1 ū3 u1 ū4 1 0

  
 
1 0 0 0  u2 ū1 u2 ū2 u2 ū3 u2 ū4   0 0 
trB (|uihu|) =
0 0 1 0 u3 ū1 u3 ū2 u3 ū3 u3 ū4 0 1
  
u4 ū1 u4 ū2 u4 ū3 u4 ū4 0 0
u ū u ū u ū u ū 0 0
  
  1 1 1 2 1 3 1 4
0 1 0 0  u2 ū1 u2 ū2 u2 ū3 u2 ū4   1 0 
+ .
0 0 0 1 u3 ū1 u3 ū2 u3 ū3 u3 ū4 0 0
 
u4 ū1 u4 ū2 u4 ū3 u4 ū4 0 1

Using matrix multiplication and matrix addition yields

 
u1 ū1 + u2 ū2 u1 ū3 + u2 ū4
trB (|uihu|) = .
u3 ū1 + u4 ū2 u3 ū3 + u4 ū4

We see that trA (|uihu|) 6= trB (|uihu|). However

tr (trA (|uihu|)) = tr (trB (|uihu|)) , det (trA (|uihu|)) = det (trB (|uihu|)) .

Problem 5. Consider the 9 × 9 matrix (density matrix)

|uihu| = (uj ūk ), j, k = 1, . . . , 9.

Find the partial trace trC3 (|uihu|), where the basis is given by
     
1 0 0
 0  ⊗ I3 ,  1  ⊗ I3 ,  0  ⊗ I3
0 0 1

and I3 denotes the 3 × 3 unit matrix.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 146

146 Problems and Solutions

Solution 5. We have
1 0 0 0 0 0
   
0 1 0 0 0 0
0 0 1 0 0 0
   
   
1 0 0 0 0 1 0 0
   
 0  ⊗ I3 = 
0 0 0,
  1  ⊗ I3 = 
0 1 0

0 0 0 0 0 0 0 1
   
0 0 0 0 0 0
   
0 0 0 0 0 0
   
0 0 0 0 0 0
and
0 0 0
 
0 0 0
0 0 0
 
 
0 0 0 0
 
 0  ⊗ I3 = 
0 0 0.

1 0 0 0
 
1 0 0
 
0 1 0
 
0 0 1
The respective transposes of the above matrices are given by
 
1 0 0 0 0 0 0 0 0
(1 0 0) ⊗ I3 =  0 1 0 0 0 0 0 0 0
0 0 1 0 0 0 0 0 0
 
0 0 0 1 0 0 0 0 0
(0 1 0) ⊗ I3 =  0 0 0 0 1 0 0 0 0
0 0 0 0 0 1 0 0 0
 
0 0 0 0 0 0 1 0 0
(0 0 1) ⊗ I3 =  0 0 0 0 0 0 0 1 0.
0 0 0 0 0 0 0 0 1
Taking this basis we find
trA (|uihu|) =
 
u1 ū1 + u4 ū4 + u7 ū7 u1 ū2 + u4 ū5 + u7 ū8 u1 ū3 + u4 ū6 + u7 ū9
 u2 ū1 + u5 ū4 + u8 ū7 u2 ū2 + u5 ū5 + u8 ū8 u2 ū3 + u5 ū6 + u8 ū9  .
u3 ū1 + u6 ū4 + u9 ū7 u3 ū2 + u6 ū5 + u9 ū8 u3 ū3 + u6 ū6 + u9 ū9

Problem 6. (i) Consider the Bell state

1
|ψi = √ (|00i + |11i).
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 147

Trace and Partial Trace 147

√
Hence c00 = c11 = 1/ 2, c01 = c10 = 0. Find ρA .
(ii) Under a unitary transformation U , V (U and V are n×n unitary matri-
ces) the matrix Λ is changed to Λ → U T ΛV , where T denotes the transpose.
Apply the transformation to ΛΛ† . Calculate the Wigner function tr(ΛΛ† )2 .

√
Solution 6. (i) Since c00 = c11 = 1/ 2, c01 = c10 = 0, we find the matrix
!
√1 0
2
Λ= .
0 √1
2

Thus we obtain the density matrix (mixed state)

1 
† 2 0
ρA = ΛΛ = .
0 12

For the other three Bell states we find the same result.
(ii) We have

ΛΛ† → (U T ΛV )(V † Λ† U T † ) = U † ΛΛ† U T †

since V † V = V V † = In . Furthermore, tr(ΛΛ† )2 stays invariant under the

transformation since U T U T † = (U † U )T = In .

Problem 7. Let { |0i, |1i, . . . , |d − 1i } be an orthonormal basis in the

Hilbert space Cd . The discrete Wigner operator is defined as
d−1 X
d−1  
X 2π
Â(q, p) := δ2q,r+s exp i p(r − s) |rihs|
r=0 s=0
d

where q and p take integer values from 0 to d − 1 and δm,u denotes the
Kronecker delta. The arithmetic in the subscript is modulo N arithmetic,
i.e. 2q mod d and (r + s) mod d. The (p, q) pairs constitute the discrete
phase space. For a state described by the density matrix ρ the discrete
Wigner function is defined as
1
W (p, q) := tr(ρÂ).
d
Let ρ = |0ih0|. Calculate W (p, q).

Solution 7. Since h0|ri = δ0r we obtain

d−1   !
1 X 2π
W (p, q) = tr |0i δ2q,s exp −i ps hs| .
d s=0
N
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 148

148 Problems and Solutions

To calculate the trace we have

d−1 d−1   !
1X X 2π
W (p, q) = hk|0i δ2q,s exp −i ps hs|ki .
d s=0
N
k=0

Using hk|0i = δk0 and hs|ki = δsk we arrive at

1
W (p, q) = δ2q,0 .
d

Problem 8. For a bipartite state with subsystems 1 and 2 described by

the joint density matrix the joint Wigner function is given by
1
W (q1 , q2 , p1 , p2 ) := tr(ρ(12) (Â1 (q1 , p1 ) ⊗ Â2 (q2 , p2 )))
d2
where the Wigner operators are given by
d−1 X
d−1  
X 2π
Â1 (q1 , p1 ) := δ2q1 ,r+s exp i p1 (r − s) |rihs|
r=0 s=0
d

and
d−1 X
d−1  
X 2π
Â2 (q2 , p2 ) := δ2q2 ,r+s exp i p2 (r − s) |rihs|.
r=0 s=0
d
Wigner functions describing a subsystem are obtained by summing the joint
Wigner functions in the corresponding set of the respective variables, e.g.
d−1 X
X d−1
W (q1 , p1 ) = W (q1 , p1 , q2 , p2 )
q2 =0 p2 =0

d−1 X
X d−1
W (q2 , p2 ) = W (q1 , p1 , q2 , p2 ).
q1 =0 p1 =0

Consider the EPR state

d−1
1 X
|ψi = √ |ki ⊗ |ki.
d k=0

Let ρ = |ψihψ|. Find W (q1 , q2 , p1 , p2 ). Discuss.

Solution 8. Straightforward calculation yields the Wigner function

1
W (q1 , q2 , p1 , p2 ) = δq ,q δp ,−p .
d2 1 2 1 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 149

Trace and Partial Trace 149

The Wigner function given above shows the connection with the EPR state
for continuous-variable teleportation

δ(q1 − q2 ) ⊗ δ(p1 + p2 )

where δ denotes the Dirac delta function.

Problem 9. The trace of an m × m matrix A over C is defined as

m
X
trm (A) = (b∗m,j A bm,j )
j=1

where { bm,1 , bm,2 , . . . , bm,m } is an orthonormal basis for Cm . The partial

traces of an mn × mn matrix C over C are defined as
m
X
tr1m,n (C) = b∗m,j ⊗ In C (bm,j ⊗ In )


j=1

n
X
tr2m,n (C) = Im ⊗ b∗n,k C (Im ⊗ bn,k )

k=1

where { bn,1 , bn,2 , . . . , bn,n } is an orthonormal basis for Cn and Im and

In denote the m × m and n × n identity matrices respectively. Let B be an
n × n matrix over C.
(i) Show that the above definition of trm (A) is independent of the choice
of orthonormal basis { bm,1 , bm,2 , . . . , bm,m }.
(ii) Show that

trmn (C) = trn tr1m,n (C) trmn (C) = trm tr2m,n (C) .



and

(iii) Use your favourite orthonormal basis for C2 to calculate

0 1 1 0
 
   
0 1 1 2 2 1 0 0 1 
tr12,2 ⊗ and tr2,2  .
1 0 3 4 1 0 0 −1
0 1 −1 0

Solution 9. (i) Let { φ1 , φ2 , . . . , φm } be an orthonormal basis for Cn .

Expanding each of the bm,k in terms of the φj and vice versa yields
m
X m
X
bm,j = φ∗k bm,j φk , φj = b∗m,k φj bm,k
k=1 k=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 150

150 Problems and Solutions

where we require
m
X m
X
b∗m,j bm,k = b∗m,j φu φ∗u φ∗v bm,k φv = b∗m,j φu φ∗v bm,k (φ∗u φv )
u,v=1 u,v=1
Xm
= φ∗u bm,k b∗m,j φu = δj,k
u=1

so that orthonormality holds. It follows that

m
X m X
X m X
m
φ∗k Aφk = φ∗k bm,j b∗m,j Ab∗m,l φk bm,l
k=1 k=1 j=1 l=1
m X
m m
!
X X
= φ∗k bm,j b∗m,l φk b∗m,j Abm,l
j=1 l=1 k=1
Xm X m
= δj,l b∗m,j Abm,l
j=1 l=1
m
X
= b∗m,j Abm,j = trm (A).
j=1

(ii) We have
n m
!
X X
trn tr1m,n (C) b∗n,j (b∗m,k


= ⊗ In )C(bm,k ⊗ In ) bn,j
j=1 k=1
Xn Xm
= (I1 ⊗ b∗n,j )(b∗m,k ⊗ In )C(bm,k ⊗ In )(I1 ⊗ bn,j )
j=1 k=1
n X
X m
= (b∗m,k ⊗ b∗n,j )C(bm,k ⊗ bn,j )
j=1 k=1
= trmn (C)
where the last equivalence follows from taking the trace of C using the basis
{ bm,j ⊗ bn,k : j ∈ {1, 2, . . . , m} k ∈ {1, 2, . . . , n} }.
The result trmn (C) = trm tr2m,n (C) follows similarly.

(iii) Using any basis, for example the standard basis

   
1 0
,
0 1
we find      
0 1 1 2 0 0
tr12,2 ⊗ =
1 0 3 4 0 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 151

Trace and Partial Trace 151

and
0 1 1 0
 
 
1 0 0 1  0 2
tr22,2  = .
1 0 0 −1 2 0
0 1 −1 0

Problem 10. Consider the GHZ state in the Hilbert space C8 (C8 ∼
=
C2 ⊗ C2 ⊗ C2 )
           
1 1 1 1 0 0 0
|GHZi = √ ⊗ ⊗ + ⊗ ⊗ .
2 0 0 0 1 1 1
Then the density matrix is given by the 8 × 8 matrix
1 0 0 0 0 0 0 1
 
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
 
10 0 0 0 0 0 0 0

ρ = |GHZihGHZ| =  .
2 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
 
0 0 0 0 0 0 0 0
 
1 0 0 0 0 0 0 1
(i) Calculate the partial trace ρAB = trC (ρ) with the basis
   
1 0
I4 ⊗ , I4 ⊗ .
0 1

(ii) Calculate the partial trace ρA = trB (ρAB ) with the basis
   
1 0
I2 ⊗ , I2 ⊗ .
0 1

Solution 10. (i) We have

  ∗      ∗   
1 1 0 0
trC (ρ) = I4 ⊗ ρ I4 ⊗ + I4 ⊗ ρ I4 ⊗
0 0 1 1
1 0 0 0 0 0 0 0
   
1 0 0 0 0 1 0 0 0 0
=  + 
2 0 0 0 0 2 0 0 0 0

0 0 0 0 0 0 0 1
1 0 0 0
 
1 0 0 0 0
=  .
2 0 0 0 0
0 0 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 152

152 Problems and Solutions

(ii) We find

ρA = trB (ρAB )
  ∗      ∗   
1 1 0 0
= I2 ⊗ ρAB I2 ⊗ + I2 ⊗ ρAB I2 ⊗
0 0 1 1
   
1 1 0 1 0 0
= +
2 0 0 2 0 1
 
1 1 0
= .
2 0 1

Programming Problems

Problem 1. Consider the hermitian matrix

 
0 0 i
A= 0 0 0.
−i 0 0

Find the eigenvalues of A utilizing

tr(A) = λ1 + λ2 + λ3 , tr(A2 ) = λ21 + λ22 + λ23 , tr(A3 ) = λ31 + λ32 + λ33 .

Solution 1. The following Maxima program will do the job

/* traceeigen.mac */
load("nchrpl");
A: matrix([0,0,%i],[0,0,0],[-%i,0,0]);
r1: mattrace(A); r2: mattrace(A . A); r3: mattrace(A . A . A);
solve([x1+x2+x3-r1=0,x1*x1+x2*x2+x3*x3-r2=0,
x1*x1*x1+x2*x2*x2+x3*x3*x3-r3=0],[x1,x2,x3]);

The eigenvalues are λ1 = −1, λ2 = 0, λ3 = +1.

Problem 2. Consider the 6 × 6 matrix A = (ajk ). Calculate the partial

trace with the basis    
1 0
⊗ I3 , ⊗ I3 .
0 1
Calculate the partial trace with the basis
     
1 0 0
 0  ⊗ I2 ,  1  ⊗ I2 ,  0  ⊗ I2 .
0 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 153

Trace and Partial Trace 153

Apply computer algebra.

Solution 2. The following Maxima program will do the job.

/* partialtrace.mac */
load("nchrpl");
A: matrix([a11,a12,a13,a14,a15,a16],[a21,a22,a23,a24,a25,a26],
[a31,a32,a33,a34,a35,a36],[a41,a42,a43,a44,a45,a46],
[a51,a52,a53,a54,a55,a46],[a61,a62,a63,a64,a65,a66]);
I2: matrix([1,0],[0,1]);
I3: matrix([1,0,0],[0,1,0],[0,0,1]);
v1: matrix([1],[0]); v2: matrix([0],[1]);
u1: matrix([1],[0],[0]); u2: matrix([0],[1],[0]);
u3: matrix([0],[0],[1]);
b1: kronecker_product(v1,I3); b1T: transpose(b1);
b2: kronecker_product(v2,I3); b2T: transpose(b2);
ptrA1: b1T . A . b1 + b2T . A . b2;
c1: kronecker_product(u1,I2); c1T: transpose(c1);
c2: kronecker_product(u2,I2); c2T: transpose(c2);
c3: kronecker_product(u3,I2); c3T: transpose(c3);
ptrA2: c1T . A . c1 + c2T . A. c2 + c3T . A. c3;

The output is the 3 × 3 matrix for the first basis

[ a44 + a11 a45 + a12 a46 + a13 ]

[ a54 + a21 a55 + a22 a46 + a23 ]
[ a64 + a31 a65 + a32 a66 + a33 ]

for the first two-dimensional basis and the 2 × 2 matrix

[ a55 + a33 + a11 a46 + a34 + a12 ]

[ a65 + a43 + a21 a66 + a44 + a22 ]

for the second three dimensional basis.

5.3 Supplementary Problems

Problem 1. Let c1 , c2 , c3 ∈ C and |c1 |2 + |c2 |2 + |c3 |2 = 1. Show that

√ √
|c1 |√2
c1 c∗2 / 2 c1 c∗2 / 2 c1 c∗3√
 
∗ 2 2 ∗
 c c / 2 |c2 | /2 |c2 | /2 c2 c3 /√2 
ρ =  1∗ 2 √
c1 c2 / 2 |c2 |2 /2 |c2 |2 /2 c2 c∗3 / 2

∗ ∗
√ ∗
√ 2
c1 c3 c2 c3 / 2 c2 c3 / 2 |c3 |
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 154

154 Problems and Solutions

is a density matrix. Taking the partial trace we obtain

√ 
(c1 c∗2 + c2 c∗3 )/ 2

|c1 |2 + |c2 |2 /2
√
ρp = .
(c∗1 c2 + c∗2 c3 )/ 2 |c3 |2 + |c2 |2 /2

Show that ρp is a density matrix. Discuss.

Problem 2. Consider the finite dimensional Hilbert spaces H1 = Cn1

and H2 = Cn2 . Let Q̂1 be an n1 × n1 matrix and Q̂2 be an n2 × n2 matrix.
Let ρ1 be a density matrix in Cn1 and ρ2 be a density matrix. Show that

tr((Q̂1 ⊗ Q̂2 )(ρ1 ⊗ ρ2 )) = tr(Q̂1 ρ1 )tr(Q2 ρ2 ).

Problem 3. Consider the finite-dimensional Hilbert spaces H1 = Cn1

and H2 = Cn2 . Let H1 ⊗ H2 be the product Hilbert space. Let |ψi and |φi
be states in the product Hilbert space H1 ⊗ H2 . Show that if

trH2 (|ψihψ|) = trH2 (|φihφ|)

then there exists a unitary matrix U acting in the Hilbert space H2 such
that
|ψi = (In1 ⊗ U )|φi
where In1 is the identity matrix in the Hilbert space H1 .

Problem 4. Consider the finite dimensional Hilbert spaces H1 = Cn1

and H2 = Cn2 . Let v, u be normalized vectors in the product Hilbert
space Cn1 ⊗ Cn2 . Show that if

tr1 (vv∗ ) = tr1 (uu∗ )

then there exists a n2 × n2 unitary matrix U such that

v = (In1 ⊗ U )u.

Problem 5. Consider the product Hilbert space `2 (N0 ) ⊗ C2s+1 , where

s = 1/2, 1, 3/2, 2, . . . is the spin. Find the partial trace over C2s+1 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 155

Chapter 6

Boolean Functions and

Quantum Gates

6.1 Introduction
A truth table (or function table) is a tabular description of a combinational
circuit (such as an AND gate, OR gate, NAND gate) listing all possible
states of the input variables together with a statement of the output vari-
able(s) for each of those possible states. The truth table for the AND gate,
OR gate, XOR gate and NOT gate are

AND OR XOR
0 0 0 0 0 0 0 0 0 NOT
0 1 0 0 1 1 0 1 1 0 1
1 0 0 1 0 1 1 0 1 1 0
1 1 1 1 1 1 1 1 0

The NAND gate is an AND gate followed by a NOT gate. The NOR gate
is an OR gate followed by a NOT gate. Both are universal gates, i.e. all
other gates can be built from these gates.

A boolean function f on n variables is a mapping {0, 1}n into {0, 1}. Let
xj ∈ {0, 1} for j = 1, . . . , n. We set x = (x1 , x2 , . . . , xn ). In the following
· denotes the AND operation, + denotes the OR operation, ⊕ the XOR

155
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 156

156 Problems and Solutions

operation and is the NOT operation. For n = 1 we have the four boolean
functions

f1 (x) = 0, f2 (x) = 1, f3 (x) = x, f4 (x) = x̄.

The last two are of course reversible.

Let X = {0, 1} and xj ∈ X. A boolean function f with n input variables,

x1 , . . . , xn and n output variables, y1 , . . . , yn is a function f : X n → X n
obeying
f (x1 , . . . , xn ) 7→ (y1 , . . . , yn ).

Here (x1 , . . . , xn ) ∈ X n is called the input vector and (y1 , . . . , yn ) ∈ X n is

called the output vector. An n-input and n-output boolean function f is
reversible if it maps each input vector to a unique output vector, i.e. the
map is a bijection.

Quantum gates are described by unitary operators. In the finite dimen-

sional Hilbert space Cd we have d × d unitary matrices. We describe
how 2n+1 × 2n+1 unitary matrices can be associated with a non-reversible
boolean function f and how 2n × 2n unitary matrices can be associated
with reversible boolean functions f . Finally the associated Hamilton oper-
ator has to be constructed.

Reversible gates are gates that function in both directions. CMOS imple-
mentations of such gates have been designed. A special pass transistor
logic family has been applied: reversible MOS. Many different reversible
logic gates are candidates as universal building blocks. The Feynman gate,
the controlled NOT gate, the Fredkin gate can be implemented. They dis-
sipate very little energy. Owing to their use of reversible truth tables, they
are even candidates for zero-power computing. Circuit synthesis takes ad-
vantage of mathematical group theory. Algorithms have been developed
for the synthesis of arbitrary reversible circuits. A reversible logic gate has
a corresponding quantum version, whose properties are completely defined
by the truth table for the classical version.

Reversible circuits are applicable to nanotechnolgy, quantum and opti-

cal computing as well as reducing power in CMOS implementations. In
adiabatic circuits, current is restricted to flow across devices with low
voltage drop and the energy stored on their capacitors is recycled. One
uses reversible energy recovery gates capable of realizing the functions
{ AN D, OR } or { N AN D, N OR }.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 157

Boolean Functions and Quantum Gates 157

6.2 Solved Problems

Problem 1. The Feynman gate is a 2 input/2 output gate given by

x01 = x1 , x02 = x1 ⊕ x2 .

(i) Give the truth table for the Feynman gate.

(ii) Show that copying can be implemented using the Feynman gate.
(iii) Show that the complement can be implemented using the Feynman
gate.
(iv) Is the Feynman gate invertible?

Solution 1. (i) The truth table is

x1 x2 x01 x02
0 0 0 0
0 1 0 1
1 0 1 1
1 1 1 0

(ii) Setting x2 = 0, we have x02 = x1 ⊕ 0 = x1 . Thus we have a copy.

(iii) Setting x2 = 1, we have x02 = x1 ⊕ 1 = x̄1 . Thus we generated the
complement.
(iv) From the truth table we see that the transformation is invertible. The
inverse transformation can be found as follows. Since x1 ⊕ x1 = 0 we have

x01 ⊕ x02 = x1 ⊕ x1 ⊕ x2 = 0 ⊕ x2 = x2 .

Thus x1 = x01 , x2 = x01 ⊕ x02 .

Problem 2. Consider the 3-input/3-output gate given by

x01 = x1 , x02 = x1 ⊕ x2 , x03 = x1 ⊕ x2 ⊕ x3 .

Give the truth table. Is the transformation invertible?

Solution 2. The truth table is given by

x1 x2 x3 x01 x02 x03
0 0 0 0 0 0
0 0 1 0 0 1
0 1 0 0 1 1
0 1 1 0 1 0
1 0 0 1 1 1
1 0 1 1 1 0
1 1 0 1 0 0
1 1 1 1 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 158

158 Problems and Solutions

From the truth table we see that the transformation is invertible, i.e. we
have a 1 − 1 map. The inverse transformation is given by

x1 = x01 , x2 = x01 ⊕ x02 , x3 = x01 ⊕ x02 ⊕ x03 .

Problem 3. Consider the 3-input/3-output gate given by

x01 = x1 , x02 = x1 ⊕ x2 , x03 = x3 ⊕ (x1 · x2 ).

Give the truth table. Is the gate invertible?

Solution 3. The truth table is given by

x1 x2 x3 x01 x02 x03

0 0 0 0 0 0
0 0 1 0 0 1
0 1 0 0 1 0
0 1 1 0 1 1
1 0 0 1 1 0
1 0 1 1 1 1
1 1 0 1 0 1
1 1 1 1 0 0

From the truth table we see that the map is invertible, i.e. we have a 1-1
map. The inverse transformation is given by

x1 = x01
x2 = x01 ⊕ x02
x3 = x03 ⊕ (x01 · (x01 ⊕ x02 ))

where we used that x ⊕ x = 0.

Problem 4. For reversible gates the following boolean expression plays

an important role
(a11 · a22 ) ⊕ (a12 · a21 )
where a11 , a12 , a21 , a22 ∈ { 0, 1 }. It could be considered as the determinant
of the 2 × 2 binary matrix
 
a11 a12
.
a21 a22

Discuss. Find the inverse of the matrix when it exists.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 159

Boolean Functions and Quantum Gates 159

Solution 4. The inverse exists iff (a11 · a22 ) ⊕ (a12 · a21 ) = 1. The inverse
is given by  
a22 a12
a21 a11
since   
a11 a12 a22 a12
=
a21 a22 a21 a11
 
(a11 · a22 ) ⊕ (a12 · a21 ) (a11 · a12 ) ⊕ (a12 · a11 )
.
(a21 · a22 ) ⊕ (a22 · a21 ) (a21 · a12 ) ⊕ (a11 · a22 )

Problem 5. Consider a two input gate (x, y) / two output gate (x0 , y 0 )
given by
x0 = a · x ⊕ b · y ⊕ c, y 0 = a0 · x ⊕ b0 · y ⊕ c0
where a, b, a0 , b0 , c, c0 ∈ { 0, 1 }.
(i) Let a = 0, b = 1, a0 = 1, b0 = 0 and c = c0 = 0. Find the output (x0 , y 0 )
for all possible inputs (x, y). Is the transformation invertible?
(ii) Let a = 1, b = 1, a0 = 1, b0 = 1 and c = c0 = 0. Find the output (x0 , y 0 )
for all possible inputs (x, y). Is the transformation invertible?

Solution 5. (i) We have

x0 = 0 · x ⊕ 1 · y ⊕ 0, y 0 = 1 · x ⊕ 0 · y ⊕ 0.

Thus
x0 = 0 ⊕ y ⊕ 0 = y, y 0 = x ⊕ 0 ⊕ 0 = x.
The truth table follows as

x y x0 y0
0 0 0 0
0 1 1 0
1 0 0 1
1 1 1 1

Therefore the transformation is invertible.

(ii) We have

x0 = 1 · x ⊕ 1 · y ⊕ 0, y 0 = 1 · x ⊕ 1 · y ⊕ 0.

Thus
x0 = x ⊕ y ⊕ 0 = x ⊕ y, y 0 = x ⊕ y ⊕ 0 = x ⊕ y.
The truth table follows as
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 160

160 Problems and Solutions

x y x0 y0
0 0 0 0
0 1 1 1
1 0 1 1
1 1 0 0

Therefore the transformation is not invertible.

Problem 6. Consider the Toffoli gate

T : {0, 1}3 → {0, 1}3 , T (a, b, c) := (a, b, (a · b) ⊕ c)

and the Fredkin gate

F : {0, 1}3 → {0, 1}3 , F (a, b, c) := (a, a · b + ā · c, a · c + ā · b)

where ā is the NOT operation, + is the OR operation, · is the AND oper-

ation and ⊕ is the XOR operation.

1. Express N OT (a) exclusively in terms of the TOFFOLI gate.

2. Express N OT (a) exclusively in terms of the FREDKIN gate.

3. Express AN D(a, b) exclusively in terms of the TOFFOLI gate.

4. Express AN D(a, b) exclusively in terms of the FREDKIN gate.

5. Express OR(a, b) exclusively in terms of the TOFFOLI gate.

6. Express OR(a, b) exclusively in terms of the FREDKIN gate.

7. Show that the TOFFOLI gate is invertible.

8. Show that the FREDKIN gate is invertible.

Thus the TOFFOLI and FREDKIN gates are each universal and reversible
(invertible).

Solution 6. (1) We have N OT (a) = BIT 3(T (a, a, 1)), where BIT 3(a, b, c) =
c. This follows from T (a, a, 1) = (a, a, a ⊕ 1).

(2) N OT (a) = BIT 3(F (a, 1, 0)).

(3) AN D(a, b) = BIT 3(T (a, b, 0)).

(4) AN D(a, b) = BIT 3(F (a, 0, b)).

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 161

Boolean Functions and Quantum Gates 161

(5) We can use

OR(a, b) = N OT (AN D(N OT (a), N OT (b)))
and simply expand N OT and AN D in terms of TOFFOLI gate. Noting
that
T (ā, b̄, 1) = (ā, b̄, (ā · b̄) ⊕ 1) = (ā, b̄, ā · b̄ = (ā, b̄, a + b)
we implement the OR operation as
OR(a, b) = BIT 3(T (N OT (a), N OT (b), 1))
= BIT 3(T (BIT 3(T (a, a, 1)), BIT 3(T (b, b, 1)), 1)).
(6) We can use
OR(a, b) = N OT (AN D(N OT (a), N OT (b)))
and simply expand N OT and AN D in terms of the FREDKIN gate. Noting
that
F (a, b, 1) = (a, a · b + ā · 1, a · 1 + ā · b) = (a, b + ā, a + b)
we implement the OR operation as
OR(a, b) = BIT 3(F (a, b, 1)).
(7) We have T −1 (a, b, c) = T (a, b, c) since the XOR (⊕) is invertible if we
remember at least one of the arguments. In other words
T (T (a, b, c)) = T (a, b, (a · b) ⊕ c)
= (a, b, (a · b) ⊕ ((a · b) ⊕ c))
= (a, b, ((a · b) ⊕ (a · b)) ⊕ c)
= (a, b, c).
(8) We have F −1 (a, b, c) = F (a, b, c) since the swap operation on two bits is
its own inverse, and the FREDKIN gate swaps the last two bits whenever
the first argument is 0. In other words
F (F (a, b, c)) = F (a, a · b + ā · c, a · c + ā · b)
= (a, a · (a · b + ā · c) + ā · (a · c + ā · b), a · (a · c + ā · b)
+ā · (a · b + ā · c))
= (a, a · b + ā · b, a · c + ā · c)
= (a, b, c).

Problem 7. A generalized Toffoli gate T(x1 , x2 , . . . , xn , xn+1 ) is a gate

that maps a boolean pattern (x1 , x2 , . . . , xn , xn+1 ) to
(x1 , x2 , . . . , xn , xn+1 ⊕ (x1 · x2 · . . . · xn ))
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 162

162 Problems and Solutions

where ⊕ is the XOR operation and · the AND operation. Show that the
generalized Toffoli gate includes the NOT gate, CNOT gate and the original
Toffoli gate.

Solution 7. The NOT gate is given by T(1, a), where a ∈ { 0, 1 }. The

CNOT-gate is given by T(a, b), where a, b ∈ { 0, 1 }. The original Toffoli
gate is given by T(a, b, c).

Problem 8. The Fredkin gate F(x1 , x2 , x3 ) has 3 inputs (x1 , x2 , x3 ) and

three outputs (y1 , y2 , y3 ). It maps boolean patterns
(x1 , x2 , x3 ) → (x1 , x3 , x2 )
if and only if x1 = 1, otherwise it passes the boolean pattern unchanged.
Give the truth table.

Solution 8. We have
x1 x2 x3 y1 y2 y3
0 0 0 0 0 0
0 0 1 0 0 1
0 1 0 0 1 0
0 1 1 0 1 1
1 0 0 1 0 0
1 0 1 1 1 0
1 1 0 1 0 1
1 1 1 1 1 1

Problem 9. The generalized Fredkin gate F(x1 , x2 , . . . , xn , xn+1 , xn+2 ) is

the mapping of the boolean pattern
(x1 , x2 , . . . , xn , xn+1 , xn+2 ) → (x1 , x2 , . . . , xn , xn+2 , xn+1 )
if and only if the boolean product x1 · x2 · . . . · xn = 1 (· is the bitwise AND
operation), otherwise the boolean pattern passes unchanged. Let n = 2
and (x1 , x2 , x3 , x4 ) = (1, 1, 0, 1). Find the output.

Solution 9. Since x1 · x2 = 1, we find

x1 x2 x3 x4 y1 y2 y3 y4
1 1 0 1 1 1 1 0

Problem 10. Is the gate (a, b, c ∈ { 0, 1 })

(a, b, c) → (a, a · b ⊕ c, a · c ⊕ b)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 163

Boolean Functions and Quantum Gates 163

reversible?

Solution 10. We set x = a, y = a · b ⊕ c and z = a · c ⊕ b. We find the

truth table
a b c x y z
0 0 0 0 0 0
0 0 1 0 1 1
0 1 0 0 0 1
0 1 1 0 1 0
1 0 0 1 0 1
1 0 1 1 1 1
1 1 0 1 1 0
1 1 1 1 0 0
Thus from the truth table we see that the gate is reversible.

Problem 11. Show that one Fredkin gate

(a, b, c) → (a, a · b + a · c, a · c + a · b)
is sufficient to implement the XOR gate. Assume that either b or c are
available.

Solution 11. Choosing b = c (equivalently c = b) we find that the Fredkin

gate yields
(a, b, c) → (a, a · b + a · b, a · c + a · c) ≡ (a, a ⊕ b, a ⊕ c).
Thus we can apply the Fredkin gate to (a, b, b) and use the second bit to
obtain a ⊕ b or equivalently apply the Fredkin gate to (a, c, c) and use the
third bit to obtain a ⊕ c.

Problem 12. Consider the 2 × 2 identity matrix and the Pauli spin
matrices σ1 , σ2 , σ3 using the following notation
 
1 0
σ00 = τ00 = I2 = ,
0 1
 
0 1
σ01 = τ01 = σ1 = ,
1 0
 
1 0
σ10 = τ10 = σ3 = ,
0 −1
 
0 −i
σ11 = σ2 = ,
i 0
 
0 1
τ11 = iσ2 = .
−1 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 164

164 Problems and Solutions

Let n be a positive integer. If v, w ∈ Zn2 with

v1 w1
   
. .
v =  ..  , w =  .. 
vn wn
and  
v
b := ∈ Z2n
2
w
we define
σb := σv1 w1 ⊗ σv2 w2 ⊗ · · · ⊗ σvn wn
and
τb := τv1 w1 ⊗ τv2 w2 ⊗ · · · ⊗ τvn wn .
Thus we can associate a bit string b with each σb and vice versa.
(i) Let n = 3 and

σb1 = σ1 ⊗ σ3 ⊗ σ1 , σb2 = I2 ⊗ σ3 ⊗ σ1 .

Find the corresponding bit strings b1 and b2 for the given σb1 and σb2 .
Then XOR the two bit strings and find the corresponding σb3 . Calculate
the matrix product σb1 σb2 . Discuss.

Solution 12. (i) We obtain the vectors

0 0
   
1 1
0 0
   
b1 =   , b2 =   .
1 0
0 0
   
1 1
The XOR operation provides the bit string
0
 
0
0
 
b3 =  
1
0
 
0
with the corresponding 8 × 8 matrix

σb3 = σ1 ⊗ I2 ⊗ I2

which is σb3 = σb1 σb2 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 165

Boolean Functions and Quantum Gates 165

Problem 13. Let x ∈ {0, 1} and |0i, |1i be the standard basis in C2 .
Consider the boolean function f (x) = x. Find a 4 × 4 permutation matrix
P such that
P
|xi ⊗ |0i → |xi ⊗ |f (x)i ≡ |xi ⊗ |xi.

Solution 13. We have to satisfy

1 0
   
0
  1
P (|0i ⊗ |0i) = |0i ⊗ |1i → P   =  
0 0
0 0

0 0
   
0 0
P (|1i ⊗ |0i) = |1i ⊗ |1i → P   =   .
1 0
0 1
A solution is
0 0 0 1
 
1 0 0 0
P = .
0 1 0 0
0 0 1 0

Problem 14. Consider the reversible gate (Feynman gate)

x1 → x1 , x2 → x1 ⊕ x2 .

The inverse function is given by (x1 , x2 ) → (x1 , x1 ⊕ x2 ). Let |0i, |1i be

the standard basis in the Hilbert space C2 . Find the unitary matrix which
implements (x1 , x2 ∈ {0, 1})

|x1 i ⊗ |x2 i 7→ |x1 i ⊗ |x1 ⊕ x2 i.

Solution 14. We have

|0i ⊗ |0i 7→ |0i ⊗ |0i, |0i ⊗ |1i 7→ |0i ⊗ |1i

|1i ⊗ |0i 7→ |1i ⊗ |1i, |1i ⊗ |1i 7→ |1i ⊗ |0i.

This provides the 4 × 4 permutation matrix

1 0 0 0
 
   
0 1 0 0 1 0 0 1
U = ≡ ⊕
0 0 0 1 0 1 1 0

0 0 1 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 166

166 Problems and Solutions

which is the CNOT-gate and ⊕ denotes the direct sum.

Problem 15. Let x1 , x2 ∈ {0, 1} and ⊕ be the XOR operation. Then

(x1 , x2 ) 7→ (x1 ⊕ 1, x1 ⊕ x2 )

is a 2-bit reversible gate since

(0, 0) 7→ (1, 0), (0, 1) 7→ (1, 1), (1, 0) 7→ (0, 1), (1, 1) 7→ (0, 0).

Let |0i, |1i be the standard basis in C2 . Find the 4 × 4 permutation matrix
P such that
P (|x1 i ⊗ |x2 i) = |x1 ⊕ 1i ⊗ |x1 ⊕ x2 i.

Solution 15. We calculate the Kronecker products of the vectors. This

provides the four equations for P

1 0 0 0
       
0 0 1 0
P   =  , P   =  ,
0 1 0 0
0 0 0 1

0 0 0 1
       
0 1 0 0
P   =  , P   =  .
1 0 0 0
0 0 1 0
Consequently we obtain the 4 × 4 permutation matrix

0 0 0 1
 
0 0 1 0
P =
1 0 0 0

0 1 0 0

with the eigenvalues +1, −1, +i, −i.

Problem 16. Given the 4 × 4 permutation matrix

0 1 0 0
 
0 0 0 1
U =
1 0 0 0

0 0 1 0

with the eigenvalues +1, −1, +i, −i. Find the corresponding boolean
function.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 167

Boolean Functions and Quantum Gates 167

Solution 16. Since the matrix has 4 = 22 rows, the function f : {0, 1}2 →
{0, 1}2 has two arguments. The first column (i.e. the column numbered 0)
has a 1 in the third row (the row numbered 2) for which b−1 (0) = (0, 0)
and b−1 (2) = (1, 0). Thus (0, 0) → (1, 0). From the second column (0, 1) →
(0, 0). The third column provides (1, 0) → (1, 1) and the fourth column
(1, 1) → (0, 1). Thus we have the map

(0, 0) 7→ (1, 0), (0, 1) 7→ (0, 0), (1, 0) 7→ (1, 1), (1, 1) 7→ (0, 1).

The right hand side provides the boolean expression

f (x1 , x2 ) = (x1 · x2 + x1 · x2 , x1 · x2 + x1 · x2 ) = (x2 , x1 ).

Problem 17. Given the boolean function f (x1 , x2 ) = x1 · x̄2 . Thus the
map is (x1 , x2 , y ∈ {0, 1})

|x1 i ⊗ |x2 i ⊗ |yi 7→ |x1 i ⊗ |x2 i ⊗ |y ⊕ (x1 · x̄2 )i

with

|0i ⊗ |0i ⊗ |0i 7→ |0i ⊗ |0i ⊗ |0i, |0i ⊗ |0i ⊗ |1i → |0i ⊗ |0i ⊗ |1i

|0i ⊗ |1i ⊗ |0i 7→ |0i ⊗ |1i ⊗ |0i, |0i ⊗ |1i ⊗ |1i 7→ |0i ⊗ |1i ⊗ |1i
|1i ⊗ |0i ⊗ |0i 7→ |1i ⊗ |0i ⊗ |1i, |1i ⊗ |0i ⊗ |1i 7→ |1i ⊗ |0i ⊗ |0i
|1i ⊗ |1i ⊗ |0i 7→ |1i ⊗ |1i ⊗ |0i, |1i ⊗ |1i ⊗ |1i 7→ |1i ⊗ |1i ⊗ |1i.
Find the 8 × 8 permutation matrix for this map.

Solution 17. The map leads to the 8 × 8 permutation matrix

 
0 1
Uf = I4 ⊕ ⊕ I2
1 0

where ⊕ denotes the direct sum and In is the n × n identity matrix.

Programming Problems

Problem 1. Consider the reversible 3-input/3-output gate given by

x01 = x1 ⊕ x3
x02 = x1 ⊕ x2
x03 = (x1 · x2 ) ⊕ (x1 · x3 ) ⊕ (x2 · x3 ).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 168

168 Problems and Solutions

The inverse is given by

x1 = x01 · x02 · x03 + x01 · x03 + x02 · x03
x2 = x01 · x02 · x03 + x01 · x03 + x02 · x03
x3 = x01 · x02 · x03 + x01 · x03 + x02 · x03 .
Give an implementation in C++ utilizing the bitset class.

Solution 1. In the bitset class & is the AND operation, | the OR opera-
tion, ^ the XOR operation and ~ the NOT operation.
// reversiblegate.cpp
#include <iostream>
#include <string>
#include <bitset>
using namespace std;

int main(void)
{
bitset<1> x1(string("1"));
bitset<1> x2(string("0"));
bitset<1> x3(string("1"));
cout << "x1 = " << x1 << endl;
cout << "x2 = " << x2 << endl;
cout << "x3 = " << x3 << endl;
bitset<1> x1p; bitset<1> x2p; bitset<1> x3p;
x1p = x1^x3; x2p = x1^x2;
x3p = (x1 & x2)^(x1 & x3)^(x2 & x3);
cout << "x1p = " << x1p << endl;
cout << "x2p = " << x2p << endl;
cout << "x3p = " << x3p << endl;
x1 = (x1p & x2p & (~x3p)) | ((~x1p) & x3p) | ((~x2p) & x3p);
x2 = ((~x1p) & x2p & (~x3p)) | (x1p & x3p) | ((~x2p) & x3p);
x3 = (x1p & (~x2p) & (~x3p)) | (~x1p & x3p) | (x2p & x3p);
cout << "x1 = " << x1 << endl;
cout << "x2 = " << x2 << endl;
cout << "x3 = " << x3 << endl;
return 0;
}

Problem 2. Given the boolean function f (x1 , x2 ) = x1 · x̄2 . Thus the

map is (x1 , x2 , y ∈ {0, 1})
|x1 i ⊗ |x2 i ⊗ |yi 7→ |x1 i ⊗ |x2 i ⊗ |y ⊕ (x1 · x̄2 )i
with
|0i ⊗ |0i ⊗ |0i 7→ |0i ⊗ |0i ⊗ |0i, |0i ⊗ |0i ⊗ |1i → |0i ⊗ |0i ⊗ |1i
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 169

Boolean Functions and Quantum Gates 169

|0i ⊗ |1i ⊗ |0i 7→ |0i ⊗ |1i ⊗ |0i, |0i ⊗ |1i ⊗ |1i 7→ |0i ⊗ |1i ⊗ |1i
|1i ⊗ |0i ⊗ |0i 7→ |1i ⊗ |0i ⊗ |1i, |1i ⊗ |0i ⊗ |1i 7→ |1i ⊗ |0i ⊗ |0i
|1i ⊗ |1i ⊗ |0i 7→ |1i ⊗ |1i ⊗ |0i, |1i ⊗ |1i ⊗ |1i 7→ |1i ⊗ |1i ⊗ |1i.
This leads to the 8 × 8 permutation matrix
 
0 1
Uf = I4 ⊕ ⊕ I2
1 0

where ⊕ denotes the direct sum and In is the n × n identity matrix. Give
a C++ implementation of this map.

Solution 2. The function main first finds the permutation matrix imple-
menting the example: f (x1 , x2 ) = x1 · x2 . Then the map (truth table) is
printed. Finally we recreate the map from the permutation matrix, which
is the reversible map

g(x1 , x2 , x3 ) = (x1 , x2 , (x1 · x2 ) ⊕ x3 ) = (x1 , x2 , x1 · x2 · x3 + x1 · x3 + x2 · x3 ).

The program counts from 0, i.e. x0 . The C++ program is

// quantumgates.cpp
#include <bitset>
#include <iostream>
#include <list>
#include <map>
#include <vector>
#include "symbolicc++.h"
using namespace std;

const int n=3;

// a class to provide ordering of bitsets

// so that they can be used in maps
template <const size_t n> class cmpbst
{
public:
bool operator()(const bitset<n> &b1,const bitset<n> &b2)
{
size_t i;
for(i=0;i<n;++i) if(b1[i] != b2[i]) return (b1[i] < b2[i]);
return false;
}
};

// for a given reversible boolean map, find the corresponding

// permutation matrix
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 170

170 Problems and Solutions

template <const size_t n>

Symbolic permutation(const map<bitset<n>,bitset<n>,cmpbst<n> > &m)
{
unsigned int N = (1 << n);
Symbolic P = Symbolic("P",N,N)*0;
typename map<bitset<n>, bitset<n> >::const_iterator i;
for(i=m.begin();i!=m.end();++i)
P(i->second.to_ulong(),i->first.to_ulong()) = 1;
return P;
}

// simplifies a sum of products form using resolution the

// products are represented by bitsets and the sum is the
// list of bitsets
template <const size_t n>
list<pair<bitset<n>,bitset<n> > > simplify(const list<bitset<n> > &s)
{
bool change = true;
// a list which indicates whether bitsets were used in resolution
// or need to be copied to the next round
list<bool> copy;
list<bool>::iterator ci1, ci2;
// each bitset is stored with a mask which indicates which bits
// may be used for resolution, once a bit is used it will be masked
list<pair<bitset<n>,bitset<n> > > r, t1, t2, *tp1=&t1, *tp2=&t2, *tpp;
typename list<bitset<n> >::const_iterator li;
typename list<pair<bitset<n>, bitset<n> > >::const_iterator lpi1, lpi2;
for(li=s.begin();li!=s.end();++li)
{
t1.push_back(make_pair(*li,bitset<n>()));
// initially all bitsets propagate
copy.push_back(true);
}
while(!tp1->empty())
{
// track whether resolution has been applied
// if no change is recorded, we are done
change = false;
for(lpi1=tp1->begin(),ci1=copy.begin();lpi1!=tp1->end();++lpi1,++ci1)
{
// search for a second bitset which differs from this bitset
// in exactly one place (taking into account the masks)
for(lpi2=lpi1,ci2=ci1;lpi2!=tp1->end();++lpi2,++ci2)
{
// only compare if the masks are the same
if(lpi1->second==lpi2->second)
{
// XOR finds the differing bits which are then masked
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 171

Boolean Functions and Quantum Gates 171

bitset<n> diff = ((lpi1->first ^ lpi2->first) & ~lpi1->second);

// only one bit differs so apply resolution
if(diff.count()==1)
{
// mask the bit which has been used
tp2->push_back(make_pair(lpi1->first,lpi1->second | diff));
change = true;
// these bitsets have been used in resolution, don’t copy them
*ci1 = *ci2 = false;
}
}
}
if(*ci1) r.push_back(*lpi1);
}
// reset the variables for the next application of resolution
tpp = tp1; tp1 = tp2; tp2 = tpp; tp2->clear();
copy.clear(); copy.resize(tp1->size(),true);
}
r.unique();
return r;
}

// find a symbolic expression for a given boolean map

template <const size_t n>
Symbolic expression(const map<bitset<n>,bitset<n>,cmpbst<n> > &m)
{
size_t j, k;
Symbolic S("S",n), NOT("NOT"), x("x",n);
vector<list<bitset<n> > > terms(n);
vector<list<pair<bitset<n>, bitset<n> > > > simplified(n);
typename map<bitset<n>,bitset<n> >::const_iterator i;
typename list<pair<bitset<n>,bitset<n> > >::iterator li;
// for each y_j, record all values of x_1,...,x_n such that y_j=1
for(i=m.begin();i!=m.end();++i)
for(j=0;j<n;++j) if(i->second[j]) terms[j].push_back(i->first);
// construct each symbolic expression for y_j
for(j=0;j<n;++j)
{
S(j) = 0;
// find a smaller set of terms
simplified[j] = simplify(terms[j]);
for(li=simplified[j].begin();li!=simplified[j].end();++li)
{
Symbolic P = 1;
for(k=0;k<n;++k)
if(!li->second[k])
{
// this is the usual construction of a product for the
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 172

172 Problems and Solutions

// sum of products form generated from a truth table

if(li->first[k]) P *= x(k); else P *= NOT[x(k)];
}
S(j) += P;
}
}
return S;
}

// determine the reversible boolean map from a permutation matrix

template <const size_t n>
map<bitset<n>,bitset<n>,cmpbst<n> > booleanmap(const Symbolic &permutation)
{
size_t i, j;
map<bitset<n>, bitset<n>, cmpbst<n> > m;
for(i=0;i<(1<<n);++i)
for(j=0;j<(1<<n);++j)
if(permutation(i,j)!=0) m[bitset<n>(j)] = bitset<n>(i);
return m;
}

// reverse the contents of a bitset

template <const int n> bitset<n> reverse(const bitset<n> &b)
{
size_t i;
bitset<n> r;
for(i=0;i<n;++i) r[n-i-1] = b[i];
return r;
}

int main(void)
{
int i1, i2, i3;
bitset<3> a, b;
map<bitset<3>, bitset<3>, cmpbst<3> > f, g;
map<bitset<3>, bitset<3>, cmpbst<3> >::const_iterator i;
Symbolic P;
for(i1=0;i1<2;++i1)
for(i2=0;i2<2;++i2)
for(i3=0;i3<2;++i3)
{
a[0] = b[0] = i1; a[1] = b[1] = i2; a[2] = i3;
b[2] = a[2]^(a[0] & (!a[1]));
f[a] = b;
}
P = permutation(f); cout << P << endl;
g = booleanmap<3>(P);
for(i=g.begin();i!=g.end();++i)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 173

Boolean Functions and Quantum Gates 173

cout << reverse<3>(i->first) << " -> " << reverse<3>(i->second) << endl;
cout << expression(f) << endl;
return 0;
}
The output is
[1 0 0 0 0 0 0 0]
[0 0 0 0 0 1 0 0]
[0 0 1 0 0 0 0 0]
[0 0 0 1 0 0 0 0]
[0 0 0 0 1 0 0 0]
[0 1 0 0 0 0 0 0]
[0 0 0 0 0 0 1 0]
[0 0 0 0 0 0 0 1]
000 -> 000
001 -> 001
010 -> 010
011 -> 011
100 -> 101
101 -> 100
110 -> 110
111 -> 111
[ x0 ]
[ x1 ]
[x0*NOT[x1]*NOT[x2]+NOT[x0]*x2+x1*x2]

6.3 Supplementary Problems

Problem 1. Given the reversible gate as truth table

x1 x2 x3 f1 f2 f3
0 0 0 0 0 1
0 0 1 0 1 0
0 1 0 0 1 1
0 1 1 1 0 0
1 0 0 1 0 1
1 0 1 1 1 0
1 1 0 1 1 1
1 1 1 0 0 0
Show that

f1 = x1 · x2 · x3 + x1 · x2 · x3 + x1 · x2 · x3 + x1 · x2 · x3
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 174

174 Problems and Solutions

f2 = x1 · x2 · x3 + x1 · x2 · x3 + x1 · x2 · x3 + x1 · x2 · x3
f3 = x1 · x2 · x3 + x1 · x2 · x3 + x1 · x2 · x3 + x1 · x2 · x3 .

Problem 2. Let x1 , x2 ∈ {0, 1} and ⊕ be the XOR operation. Then

(x1 , x2 ) 7→ (x1 ⊕ 1, x1 ⊕ x2 )

is a 2-bit reversible gate since

(0, 0) 7→ (1, 0), (0, 1) 7→ (1, 1), (1, 0) 7→ (0, 1), (1, 1) 7→ (0, 0).

Let |0i, |1i be the standard basis in C2 . Find the 4 × 4 permutation matrix
P such that
P (|x1 i ⊗ |x2 i) = |x1 ⊕ 1i ⊗ |x1 ⊕ x2 i.

Problem 3. The Toffoli gate T(x1 , x2 , x3 ) has 3 inputs (x1 , x2 , x3 ) and

three outputs (y1 , y2 , y3 ) and is given by

(x1 , x2 , x3 ) → (x1 , x2 , x3 ⊕ (x1 · x2 ))

where x1 , x2 , x3 ∈ { 0, 1 }, ⊕ is the XOR operation and · the AND operation.

Show that the truth table is given by

x1 x2 x3 y1 y2 y3
0 0 0 0 0 0
0 0 1 0 0 1
0 1 0 0 1 0
0 1 1 0 1 1
1 0 0 1 0 0
1 0 1 1 0 1
1 1 0 1 1 1
1 1 1 1 1 0

Problem 4. Consider the 3-input/3-output gate given by

x01 = x1 ⊕ x3 , x02 = x1 ⊕ x2 , x03 = (x1 · x2 ) ⊕ (x1 · x3 ) ⊕ (x2 · x3 ).

Give the truth table. Show that the gate is invertible. Find the inverse
function.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 175

Chapter 7

Unitary Transforms and

Quantum Gates

7.1 Introduction
Quantum gates are realised as unitary operators. Let H denote a Hilbert
space. A linear operator U in a Hilbert space is unitary if

U ∗ = U −1 .

In other words, U is unitary if and only if U is invertible and U −1 = U ∗ .

An equivalent form of definition is: U is unitary iff

hU x, U yi = hx, yi, for all x, y ∈ H .

Thus we have
U ∗U = U U ∗ = I
where ∗ denotes the adjoint and I is the identity operator. The composite
of two unitary operators is again unitary.

Here we consider n × n unitary matrices. If U1 , U2 are n × n unitary

matrices, then U1 U2 is an n × n unitary matrix. The n × n unitary matrices
form a group under matrix multiplication. All the eigenvalues of a unitary
matrix have absolute value equal to 1, i.e. |λ| = 1. Thus | det(U )| = 1.

175
January 30, 2018 15:21 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 176

176 Problems and Solutions

All n × n unitary matrices form a group under matrix multiplication. An

important subgroup are all n × n unitary matrices with det(U ) = 1. If
det(U1 ) = 1 and det(U2 ) = 1, then det(U1 U2 ) = 1. Another important
subgroup are all n × n permutation matrices. A number of quantum gates
are given as permutation matrices such as

1 0 0 0 1 0 0 0
   
0 1 0 0 0 0 1 0
UCN OT = , USW AP = .
0 0 0 1 0 1 0 0
0 0 1 0 0 0 0 1

If K is a hermitian matrix, then exp(iK) is a unitary matrix. Thus if Ĥ

is a hermitian matrix describing the Hamilton operator we find that the
exponential function
exp(−iĤt/~)
is a unitary matrix. Let A be an arbitrary n × n matrix over C. Then

U exp(A)U ∗ = exp(U AU ∗ ).

If U and V are unitary, then U ⊗ V and U ⊕ V are unitary, where ⊗ is the

Kronecker product and ⊕ the direct sum.
The square roots of a unitary matrix are not necessarily unitary matrices
again.
Let v be a normalized state in Cn . Then U v is a normalized state.
Let v1 , v2 , . . . , vn be an orthonormal basis in Cn . Then

V = ( v1 v2 ... vn )

is a unitary matrix. Let |λj | = 1. Then

n
X
U= λj vj vj∗
j=1

is a unitary matrix.
An important unitary matrix is the Fourier matrix given by

1 1 1 ... 1
 
1 ω ω2 ... ω n−1 
1  1 ω 2
ω 4
... ω 2(n−1) 

F =√ 
n  ..
 .. .. .. .. 
. . . . . 
n−1 2(n−1) (n−1)(n−1)
1 ω ω ... ω

where ω n = 1.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 177

Unitary Transforms and Quantum Gates 177

7.2 Solved Problems

Problem 1. Let U be a 2 × 2 unitary matrix and I2 be the 2 × 2 identity
matrix. Is the 4 × 4 matrix
   iα 
0 0 e 0
V (α) = ⊗U + ⊗ I2 , α∈R
0 1 0 0
unitary?

Solution 1. Yes, V (α) is unitary. Since

   iα   iα    
0 0 e 0 e 0 0 0 0 0
= =
0 1 0 0 0 0 0 1 0 0
and U U ∗ = I2 we obtain
   
∗ 0 0 1 0
V (α)V (α) = ⊗ I2 + ⊗ I2 = I2 ⊗ I2 .
0 1 0 0

Problem 2. (i) Let

1 i 0 0
 
1 0 0 i 1 
M := √  .
2 0 0 i −1
1 −i 0 0
Is the matrix M unitary? In quantum computing M is called the magic
gate.
(ii) Let    
1 1 1 1 0
UH := √ , US :=
2 1 −1 0 i
and
1 0 0 0
 
0 0 0 1
UCN OT 2 =  .
0 0 1 0
0 1 0 0
Show that the matrix M can be written as

M = UCN OT 2 (I2 ⊗ UH )(US ⊗ US ).

(iii) Let SO(4) be the special orthogonal Lie group. Let SU (2) be the
special unitary Lie group. Show that for every real orthogonal matrix U ∈
SO(4), the matrix M U M −1 is the Kronecker product of two 2-dimensional
special unitary matrices, i.e.

M U M −1 ∈ SU (2) ⊗ SU (2).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 178

178 Problems and Solutions

Solution 2. (i) Since M M ∗ = I4 we find that M is unitary.

(ii) We obtain
 √12 √1 0 0
 
1 0 0 0 2 1 0 0 0
  
 √1 − √1 0 0
0 0 0 1 2 0 i 0 0 

2
M = .
 
0 0 1 0  0 0 √1 √1  0 0 i 0
 
2 2
0 1 0 0 0 0 √1 1
− √2 0 0 0 −1
2

(iii) We show that for every A ⊗ B ∈ SU (2) ⊗ SU (2), we have

M −1 (A ⊗ B)M ∈ SO(4).
Now every matrix A ∈ SU (2) can be written as Rz (α)Ry (θ)Rz (β) for some
α, β, θ ∈ R, where
   −iα/2 
cos(θ/2) sin(θ/2) e 0
Ry (θ) = , Rz (α) = .
− sin(θ/2) cos(θ/2) 0 eiα/2
Therefore any matrix A ⊗ B ∈ SU (2) ⊗ SU (2) can be written as a product
of the matrices of the form V ⊗ I2 or I2 ⊗ V , where V is either Ry (θ) or
Rz (α). Next we have to show that M −1 (V ⊗ I2 )M and M −1 (I2 ⊗ V )M are
in SO(4). We have
cos(θ/2) 0 0 − sin(θ/2)
 
0 cos(θ/2) sin(θ/2) 0
M −1 (Ry (θ) ⊗ I2 )M = 
 
0 − sin(θ/2) cos(θ/2) 0

sin(θ/2) 0 0 cos(θ/2)
cos(α/2) sin(α/2) 0 0
 
 − sin(α/2) cos(α/2) 0 0
M −1 (Rz (α) ⊗ I2 )M =  .

0 0 cos(α/2) − sin(α/2)
0 0 sin(α/2) cos(α/2)
We have similar equations for the cases of I2 ⊗ Ry (θ) and I2 ⊗ Rz (α). Since
the mapping
A ⊗ B → M −1 (A ⊗ B)M
is one-to-one (invertible) and the Lie groups SU (2) ⊗ SU (2) and SO(4)
have the same topological dimension, we conclude that the mapping is an
isomorphism between these two Lie groups.

Problem 3. (i) Let A := |0ih0|−|1ih1| in the Hilbert space C2 . Calculate

UH AUH |0i, UH AUH |1i
where UH is the Walsh-Hadamard transform. The unitary transform UH is
defined by
1
UH |ki = √ (|0i + (−1)k |1i), k ∈ {0, 1}.
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 179

Unitary Transforms and Quantum Gates 179

(ii) Calculate
(UH ⊗ UH )UCN OT (UH ⊗ UH )|j, ki
where |j, ki ≡ |ji ⊗ |ki with j, k ∈ {0, 1}, and the answer is in the form of
a ket |m, ni with m, n ∈ {0, 1}. The unitary transform

UCN OT := |0ih0| ⊗ I2 + |1ih1| ⊗ UN OT

is the controlled NOT operation and the unitary transform

UN OT = |0ih1| + |1ih0|

is the NOT operation.

Solution 3. (i) Let j ∈ {0, 1}. Then

1 1
UH AUH |ji = √ UH A(|0i + (−1)j |1i) = √ UH (|0i + (−1)j+1 |1i)
2 2
1 j
= √ UH (|0i + (−1) |1i) = |ji
2
where j := 1 − j. In other words UH AUH = UN OT .
(ii) Straightforward calculation yields

(UH ⊗ UH )UCN OT (UH ⊗ UH )|j, ki

1
= (UH ⊗ UH )UCN OT (|0i + (−1)j |1i) ⊗ (|0i + (−1)k |1i)


2
1
= (UH ⊗ UH )(|00i + (−1)k |01i + (−1)j |11i + (−1)j+k |10i)
2
1
= (UH ⊗ UH )(|0i ⊗ (|0i + (−1)k |1i) + (−1)j |1i ⊗ (|1i + (−1)k |0i))
2
1
= (UH ⊗ UH )(|0i ⊗ (|0i + (−1)k |1i) + (−1)j+k |1i ⊗ (|0i + (−1)k |1i))
2
1
= (UH ⊗ UH )(|0i + (−1)j+k |1i) ⊗ (|0i + (−1)k |1i)
2
= |j ⊕ k, ki

where ⊕ is the XOR operation. In other words we have the controlled NOT
operation, where the control qubit is the second qubit and the target qubit
is the first qubit.

Problem 4. Consider the linear operator (Hamilton operator)

H := i~ω(|0ih1| − |1ih0|)

operating in the Hilbert space C2 , where { |0i, |1i } is an orthonormal basis

in C2 and ω is a real parameter (frequency).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 180

180 Problems and Solutions

(i) Is H self-adjoint?
(ii) Find the eigenvalues and corresponding normalized eigenvectors of H.
(iii) Find the unitary matrix U (t) := exp(−iHt/~). Find the values of t
such that U (t) performs the NOT operation
U (t)|0i → |1i, U (t)|1i → |0i.
(iv) Calculate U (t = π/4ω) and (U (t = π/4ω))2 .

Solution 4. (i) The adjoint of an operator can be obtained by simply

swapping the labels of the corresponding bra and ket vectors in the sum,
and taking the complex conjugate of all complex coefficients. Thus
H ∗ = i~ω(|1ih0| − |0ih1|) = −i~ω(|1ih0| − |0ih1|) = H
i.e. H is self-adjoint. We can determine H ∗ as follows. Let
H ∗ = a00 |0ih0| + a01 |0ih1| + a10 |1ih0| + a11 |1ih1|, a00 , a01 , a10 , a11 ∈ C.
The bra vector corresponding to the ket H|yi is hy|H ∗ . We require that
hH ∗ y|xi = hy|Hxi for all |xi = x0 |0i + x1 |1i and |yi = y0 |0i + y1 |1i. We
find
H|xi = i~ω(x1 |0i − x0 |1i)
H ∗ |yi = (y0 a00 + y1 a01 )|0i + (y0 a10 + y1 a11 )|1i
hy|Hxi = i~ω(x1 y0 − x0 y1 )
hH ∗ y|xi = x0 (y0 a00 + y1 a01 ) + x1 (y0 a10 + y1 a11 ).
Since hH ∗ y|xi = hy|Hxi for all |xi and |yi, we obtain
i~ωy0 = (y0 a10 + y1 a11 ), −i~ωy1 = (y0 a00 + y1 a01 ) .
Consequently a00 = 0, a01 = i~ω, a10 = −i~ω, a11 = 0.
(ii) The eigenvalue equation for H is
H(a|0i + b|1i) = λ(a|0i + b|1i).
Thus we have the two equations −i~ωa = λb, i~ωb = λa. If λ = 0 we have
a = 0 and b = 0. Therefore we only have to consider λ 6= 0. Obviously we
may assume b 6= 0 (thus a 6= 0). We obtain
i~ωa
λ=− .
b
Hence ib2 = −ia2 so that b = ±ia. Using |a|2 + |b|2 = 1 we find |a| = √12 .
We obtain the eigenvalues and corresponding orthonormal eigenvectors
1 1
λ1 = −~ω, √ (|0i + i|1i), λ2 = ~ω, √ (|0i − i|1i).
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 181

Unitary Transforms and Quantum Gates 181

(iii) We find H n (n ∈ N) by observing that

H 2 = (~ω)2 (|0ih0|+|1ih1|) = (~ω)2 I2 , H 3 = (~ω)2 H, H 4 = (~ω)4 I2 .

Thus
(~ω)n−1 H n odd

n
H = .
(~ω)n I2 n even
Since U (t) := exp(−iHt/~) we have
∞ ∞ ∞
X (− it )j H j
~
X (−iωt)2j 1 X (−iωt)2j+1
U (t) = = I2 + H
j=0
j! j=0
(2j)! ~ω j=0 (2j + 1)!
∞ ∞
X (−1)j (ωt)2j 1 X (−1)j (ωt)2j+1
= I2 − i H
j=0
(2j)! ~ω j=0 (2j + 1)!
i
= cos(ωt)I2 − sin(ωt)H
~ω
= cos(ωt)(|0ih0| + |1ih1|) + sin(ωt)(|0ih1| − |1ih0|).

For the NOT operation we use U (t = π/2ω) = |0ih1| − |1ih0|. The unitary
transforms U ((2k + 1)π/2ω), k ∈ N0 implement the NOT operation.
(iv) We have
1 1
U (t = π/4ω) = √ (|0i − |1i)h0| + √ (|0i + |1i)h1|
2 2
2
U (t = π/4ω) = U (t = π/2ω) = |0ih1| − |1ih0|.

Thus we find (U (t = π/4ω))2 = U (t = π/2ω), i.e. U (t = π/4ω) acts as the

square root of the NOT operation. Traditionally in √ quantum computation
we use UN OT = |0ih1| + |1ih0|. In this case for the NOT operation we use
1 1
U√N OT = (1 + i)(|0ih0| + |1ih1|) + (1 − i)(|0ih1| + |1ih0|).
2 2

Problem 5. Consider the Hilbert space Cn . Let e1 , e2 , . . . , en be the

standard basis in Cn , Sn be the symmetric group of order n! and Uσ be the
unitary matrix on ⊗n Cn such that

Uσ (e1 ⊗ · · · ⊗ en ) := eσ(1) ⊗ · · · ⊗ eσ(n)

where σ ∈ Sn . We define the matrix (“antisymmetrization operator”) in

the Hilbert space ⊗n Cn by
1 X
Πn := sgn(σ)Uσ
n!
σ∈Sn
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 182

182 Problems and Solutions

where sgn is the signature of the permutation σ ∈ Sn . The matrices Πn

are projection matrices. Find Π2 and Π3 .

Solution 5. (i) For n = 2 we have the unitary matrices

       
1 0 1 0 0 1 0 1
U11,22 = ⊗ , U12,21 = ⊗
0 1 0 1 1 0 1 0

and therefore
1 0 0 −1
 
1 1 0 1 −1 0 
Π2 = (U11,22 − U12,21 ) =  .
2 2 0 −1 1 0
−1 0 0 1

Problem 6. (i) Let σ1 , σ2 , σ3 be the Pauli spin matrices. Find

R1x (α) := exp(−iα(σ1 ⊗ I2 )), R1y (α) := exp(−iα(σ2 ⊗ I2 ))

where α ∈ R and I2 denotes the 2 × 2 unit matrix.

(ii) Consider the special case R1x (α = π/2) and R1y (α = π/4). Calculate
the matrix R1x (π/2)R1y (π/4). Discuss.

Solution 6. (i) We have

∞
X (−iα(σ1 ⊗ I2 ))k
exp(−iα(σ1 ⊗ I2 )) := .
k!
k=0

Since σ12 = I2 we have (σ1 ⊗ I2 )2 = I2 ⊗ I2 . Thus we find

exp(−iα(σ1 ⊗ I2 )) = (I2 ⊗ I2 ) cos α + e−iπ/2 (σ1 ⊗ I2 ) sin(α)

where we used exp(−iπ/2) = −i. Analogously, we find

exp(−iα(σ2 ⊗ I2 )) = (I2 ⊗ I2 ) cos(α) + e−iπ/2 (σ2 ⊗ I2 ) sin(α)

since (σ2 ⊗ I2 )2 = I2 ⊗ I2 .
(ii) Since sin(π/2) = 1, cos(π/2) = 0 we arrive at

R1x (π/2) = e−iπ/2 (σ1 ⊗ I2 ).

√ √
From sin(π/4) = 2/2, cos(π/4) = 2/2 it follows that
1 1
R1y (π/4) = √ (I2 ⊗ I2 ) + √ e−iπ/2 (σ2 ⊗ I2 ).
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 183

Unitary Transforms and Quantum Gates 183

Thus
e−iπ/2 e−iπ/2
R1x (π/2)R1y (π/4) = √ (σ1 ⊗ I2 ) + √ (σ3 ⊗ I2 )
2 2
where we used that σ1 σ2 = iσ3 . Therefore

e−iπ/2
R1x (π/2)R1y (π/4) = √ (σ1 + σ3 ) ⊗ I2
2
where  
1 1 1 1
√ (σ1 + σ3 ) = √
2 2 1 −1
is the Walsh-Hadamard gate. All the single operations are in the Lie
group SU (2) whose determinant is +1, while the determinant of the Walsh-
Hadamard gate is −1. Thus the overall phase is unavoidable.

n
Problem 7. Consider the Hilbert space C2 . Let { |0i, |1i, . . . , |2n −1i }
be an orthonormal basis in this Hilbert space. We define the linear operator
n n
2 −1 2 −1
1 X X −i2πkj/2n
UQF T := √ e |kihj| . (1)
2n j=0 k=0

This transform is called the quantum Fourier transform . Show that UQF T
∗
is unitary. In other words show that UQF T UQF T = I2n , where we use the
completeness relation
n
2X −1
I2n = |jihj|.
j=0
n n
Thus I2n is the 2 × 2 unit matrix.

Solution 7. From the definition (1) we find

n n
2 −1 2 −1
∗ 1 X X i2πkj/2n
UQF T =√ e |jihk|
2n j=0 k=0

∗
where denotes the adjoint. Therefore
n n n n
2 −1 2 −1 2 −1 2 −1
∗ 1 X X X X i2π(kj−lm)/2n
UQF T UQF T = n e |jihk|lihm|
2 j=0 m=0
k=0 l=0
n n n
2X −1 2X −1 2X −1
1 n
= ei2π(kj−km)/2 |jihm|.
2n j=0 k=0 m=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 184

184 Problems and Solutions

n
We have for j = m, ei2π(kj−km)/2 = 1. Thus for j, m = 0, 1, . . . , 2n − 1
n
2X −1
n
(ei2π(j−m)/2 )k = 2n , j=m
k=0
n
2X −1
n 1 − ei2π(j−m)
(ei2π(j−m)/2 )k = = 0, j 6= m.
k=0
1 − ei2π(j−m)/2n

Thus
n
2X −1
∗
UQF T UQF T = |jihj| = I2n .
j=0

Problem 8. Apply the quantum Fourier transform to the state in the

Hilbert space C8
7
1X
cos(2πj/8)|ji
2 j=0

where the quantum Fourier transform is given by

7 7
1 X X −i2πkj/8
UQF T = √ e |kihj|.
2 2 j=0 k=0

We use { |ji : j = 0, 1, . . . , 7 } as an orthonormal basis in the Hilbert space

C8 , where |7i = |111i ≡ |1i ⊗ |1i ⊗ |1i.

Solution 8. We use Euler’s identity eiθ ≡ cos(θ) + i sin(θ) and

N
X −1
ei2πk(n−m)/N = N δnm .
k=0

Thus we have
7 7
X 1 X  i2π(1−k)j/8 
x̂(k) = e−i2πkj/8 cos(2πj/8) = e + e−i2π(1+k)j/8
j=0
2 j=0
= 4(δk1 + δk7 )

and
7 7
1X 1 X 1
UQF T cos(2πj/8)|ji = √ x̂(k)|ki = √ (|1i + |7i).
2 j=0 2 8 k=0 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 185

Unitary Transforms and Quantum Gates 185

Problem 9. Let
n n
2X −1 2X −1  
2
UIA := − δjk |kihj|. (1)
j=0 k=0
2n

UIA is called the inversion about average operator. Show that UIA is uni-
∗
tary. In other words show that UIA UIA = I2n .
Hint: Use the completeness relation
n
2X −1
|jihj| = I2n .
j=0

Solution 9. From (1) we find

n n
2X −1 2X −1  
∗ 2
UIA = − δ jk |kihj| = UIA .
j=0 k=0
2n

Thus
n
2X −1   
∗ 2 2 2
UIA UIA = UIA = − δjk − δlm |kihj|mihl|
2n 2n
j,k,l,m=0
n
2X −1   
2 2
= − δjk − δlj |kihl|
2n 2n
j,k,l=0

where we used that hj|mi = δjm . Furthermore, we find

n n
2X −1    2X −1  
2 2 4 2 2
− δjk − δ lj = − δ jk − δ lj + δ jk δ lj
j=0
2n 2n j=0
22n 2n 2n
n n
2X−1 2X−1
4 2 2
= n− n− n+ δjk δlk = δlk δjk
2 2 2 j=0 j=0
= δlk .
Therefore n
2X −1
∗
UIA UIA = |jihj| = I2n .
j=0

Problem 10. Let { |0i, |1i } be an orthonormal basis in the two-dimensional

Hilbert space C2 and
1
UH |ki := √ |0i + (−1)k |1i , k ∈ {0, 1}


2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 186

186 Problems and Solutions

UP S(θ) := |00ih00| + |01ih01| + |10ih10| + eiθ |11ih11|

UCN OT := |00ih00| + |01ih01| + |10ih11| + |11ih10| .

(i) From these definitions show that UH UH = I2 .

(ii) Calculate
(I2 ⊗ UH )UP S(π) (I2 ⊗ UH )|abi
and (I2 ⊗ UH )UCN OT (I2 ⊗ UH )|abi, where a, b ∈ {0, 1}. What is the use of
these transforms?

Solution 10. (i) An arbitrary state in the Hilbert space C2 can be written
as
|ψi := a|0i + b|1i
where a, b ∈ C and |a|2 + |b|2 = 1. We find
1 1
UH UH |ψi = UH √ (a|0i + a|1i + b|0i − b|1i) = (2a|0i + 2b|1i)
2 2
= a|0i + b|1i.

Thus, UH UH = I2 .
(ii) We find
1
(I2 ⊗ UH )UP S(π) (I2 ⊗ UH )|abi = (I2 ⊗ UH )UP S(π) √ |ai ⊗ (|0i + (−1)b |1i)
2
1
= (I2 ⊗ UH ) √ |ai ⊗ (|0i + (−1)a+b |1i
2
1
= |a, a ⊕ bi
2
where a ⊕ b = a + b (modulo 2) is the XOR operation. We obtain
1
(I2 ⊗ UH )UCN OT (I2 ⊗ UH )|abi = (I2 ⊗ UH )UCN OT √ |ai ⊗ (|0i + (−1)b |1i)
2
(
1
(I2 ⊗ UH ) √2 |ai ⊗ (|0i + (−1)b |1i) a = 0
=
(I2 ⊗ UH ) √12 |ai ⊗ (|1i + (−1)b |0i) a = 1
1
= (I2 ⊗ UH ) √ |ai ⊗ (−1)ab (|0i + (−1)b |1i)
2
ab
= (−1) |abi.

The first computation is UCN OT , the second is UP S(π) .

Problem 11. The XOR gate is given by

UXOR (|mi ⊗ |ni) = |mi ⊗ |m ⊕ ni

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 187

Unitary Transforms and Quantum Gates 187

where m, n ∈ {0, 1} and ⊕ denotes addition modulo 2. The transformation

has the following properties: (a) it is unitary and thus reversible, (b) it is
hermitian, (c) m ⊕ n = 0 if and only if m = n. The first index denotes
the state of the control qubit and the second index denotes the state of the
target qubit.
A generalized quantum XOR gate (GXOR gate) acts on two d-dimensional
quantum systems (d > 2). In analogy with qubits one calls these two
systems qudits. The basis states |mi of each qudit are labeled by elements
in the ring Zd which we denote by the numbers, m = 0, 1, . . . , d − 1, with
the usual rules for addition and multiplication modulo d. We define two
operators
UGXOR1 (|mi ⊗ |ni) := |mi ⊗ |m ⊕ ni
and
UGXOR2 (|mi ⊗ |ni) := |mi ⊗ |m ni
where m n := (m − n) modulo d. Discuss the properties of these two
operators.

Solution 11. For UGXOR1 we find that the operator is unitary but not
hermitian for d > 2. Therefore it is no longer its own inverse. We have to
obtain the inverse of the UGXOR1 gate by iteration, i.e.
−1 d−1 †
UGXOR1 = UGXOR1 = UGXOR1 6= UGXOR1 .

For the operator UGXOR2 we find that in the special case for d = 2 it
reduces to the XOR-gate. Furthermore, the operator is unitary, hermitian
and m n = 0 modulo d if and only if m = n.

Problem 12. Given an orthonormal basis in CN denoted by

|φ0 i , |φ1 i , . . . , |φN −1 i.

(i) Show that

N
X −2
U := |φk ihφk+1 | + |φN −1 ihφ0 |
k=0

is a unitary matrix. Find tr(U ).

(ii) Find U N .
(iii) Does U depend on the chosen basis? Prove or disprove.
Hint. Consider N = 2, the standard basis (1, 0)T , (0, 1)T and the basis
√1 (1, 1)T , √1 (1, −1)T .
2 2
(iv) Show that the set
{ U, U 2 , . . . , U N }
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 188

188 Problems and Solutions

forms a commutative group (abelian group) under matrix multiplication.

The set is a subgroup of the group of all permutation matrices.
(v) Assume that the set given above is the standard basis. Show that the
unitary matrix U is given by the permutation matrix
0 1 0 ... 0
 
0 0 1 ... 0
. . . .. .. 
U = . . .
. . . . ..
0 0 0 ... 1
1 0 0 ... 0

Solution 12. (i) Since hφj |φk i = δjk we have

−2
N
! N −2 !
X X
∗
UU = |φk ihφk+1 | + |φN −1 ihφ0 | |φk+1 ihφk | + |φ0 ihφN −1 |
k=0 k=0
N
X −1
= |φk ihφk | = IN .
k=0

Obviously we have tr(U ) = 0 since the terms |φk ihφk | do not appear in the
sum (i.e. we calculate the trace in the basis { |φ0 i, . . . , |φN −1 i }).
(ii) We notice that U maps |φk i to |φk−1 i. Applying this N times and using
modulo N arithmetic we obtain (i.e. U N maps |φk i to |φk−N i)
U N = IN .
(iii) For the standard basis in C2 { (1, 0)T , (0, 1)T } we obtain
 
0 1
Ustd = .
1 0
For the basis in C2 { √12 (1, 1)T , √12 (1, −1)T } we obtain the Pauli spin ma-
trix σ3 . Obviously the two unitary matrices are different. Of course there
is a unitary matrix V such that Ustd = V −1 σ3 V .
(iv) Since UN = IN = U 0 we have that
U s U t = U s+t = U s+t mod N .
Thus the set of matrices {U, U 2 , . . . , U N } forms an abelian group under ma-
trix multiplication, because {0, 1, . . . , N − 1} forms a group under addition
modulo N . The two groups are isomorphic.
(v) Let ej denote the element of the standard basis in Cn with a 1 in the
jth position (numbered from 0) and 0 in all other positions. Then U is
given by
NX−2
U= ek eTk+1 + eN −1 eT0 .
k=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 189

Unitary Transforms and Quantum Gates 189

In the product ek eTk+1 , ek denotes the row and eTk+1 denotes the column in
the matrix U . Thus we obtain the matrix described above.

Problem 13. (i) Let σ1 , σ2 and σ3 be the Pauli spin matrices and I2 be
the 2 × 2 unit matrix. Find

(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 )(σ1 ⊗ σ1 ⊗ I2 ⊗ I2 )(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 ),

(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 )(I2 ⊗ σ1 ⊗ σ1 ⊗ I2 )(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 ),

(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 )(I2 ⊗ I2 ⊗ σ1 ⊗ σ1 )(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 ).
(ii) Replace σ1 by σ2 in the expressions given above and calculate the
expressions.
(iii) Given the one-dimensional XY -model with open boundary conditions

N/2−1   N/2
X 1+γ 1−γ X
ĤXY = − σ1,j σ1,j+1 + σ2,j σ2,j+1 −λ σ3,j
2 2
j=−N/2+1 j=−N/2+1

where the parameter λ is the intensity of the magnetic field applied in the
z-direction and the parameter γ determines the degree of anisotropy of the
spin-spin interaction, which is restricted to the xy-plane in spin space. Find
   
N/2 N/2
Y Y
 σ3,j  ĤXY  σ3,j  .
j=−N/2+1 j=−N/2+1

Solution 13. (i) Since σ32 = I2 and σ3 σ1 σ3 = −σ1 we find for the first
expression

(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 )(σ1 ⊗ σ1 ⊗ I2 ⊗ I2 )(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 ) = σ1 ⊗ σ1 ⊗ I2 ⊗ I2 .

Analogously, we find

(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 )(I2 ⊗ σ1 ⊗ σ1 ⊗ I2 )(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 ) = I2 ⊗ σ1 ⊗ σ1 ⊗ I2

(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 )(I2 ⊗ I2 ⊗ σ1 ⊗ σ1 )(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 ) = I2 ⊗ I2 ⊗ σ1 ⊗ σ1 .

(ii) Replacing σ1 by σ2 and using σ3 σ2 σ3 = −σ2 yields

(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 )(σ2 ⊗ σ2 ⊗ I2 ⊗ I2 )(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 ) = σ2 ⊗ σ2 ⊗ I2 ⊗ I2

(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 )(I2 ⊗ σ2 ⊗ σ2 ⊗ I2 )(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 ) = I2 ⊗ σ2 ⊗ σ2 ⊗ I2

(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 )(I2 ⊗ I2 ⊗ σ2 ⊗ σ2 )(σ3 ⊗ σ3 ⊗ σ3 ⊗ σ3 ) = I2 ⊗ I2 ⊗ σ2 ⊗ σ2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 190

190 Problems and Solutions

(iii) Using the results from (i) and (ii) and extending from N = 4 to arbi-
trary N , we find
   
N/2 N/2
Y Y
 σ3,j  ĤXY  σ3,j  = ĤXY .
j=−N/2+1 j=−N/2+1

From (ii) and (iii) we find that the Hamilton operator ĤXY is invariant
under this transformation.

Problem 14. (i) Consider the product state |Di ⊗ |P i, where |Di is
a state to describe a m-qubit data register and |P i is a state to describe
an n-qubit program register. Let G be a unitary operator acting on this
product state
|Di ⊗ |P i → G(|Di ⊗ |P i).
The unitary operator is implemented as follows. A unitary operator U
acting on the m-qubits of the data register, is said to be implemented by
this gate array if there exists a state |PU i of the program register such that

G(|Di ⊗ |PU i) = (U |Di) ⊗ |PU0 i

for all states |Di of the data register and some state |PU0 i of the program
register. Show that |PU0 i does not depend on |Di.
(ii) Suppose distinct (up to a global phase) unitary operators U1 , . . . , UN
are implemented by some programmable quantum gate array. Show that
the corresponding programs |P1 i, . . . , |PN i are mutually orthogonal.

Solution 14. (i) Consider

G(|D1 i ⊗ |P i) = (U |D1 i) ⊗ |P10 i, G(|D2 i ⊗ |P i) = (U |D2 i) ⊗ |P20 i.

Taking the scalar product of these two equations and using G† G = I,

U † U = I and hP |P i = 1 we find

hD1 |D2 i = hD1 |D2 ihP10 |P20 i.

6 0 we find hP10 |P20 i = 1. Thus |P10 i = |P20 i. Consequently, there

If hD1 |D2 i =
is no |Di dependence of |PU0 i. What happens for hD1 |D2 i = 0 ?
(ii) Suppose that |P i and |Qi are programs which implement unitary op-
erators Up and Uq which are distinct up to global phase changes. Then for
an arbitrary data state |Di we have

G(|Di ⊗ |P i) = (Up |Di) ⊗ |P 0 i

G(|Di ⊗ |Qi) = (Uq |Di) ⊗ |Q0 i

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 191

Unitary Transforms and Quantum Gates 191

where |P 0 i and |Q0 i are states of the program register. Taking the scalar
product of these two equations and using G† G = I, hD|Di = 1 we obtain

hQ|P i = hQ0 |P 0 ihD|Uq† Up |Di.

Suppose that hQ0 |P 0 i =

6 0. Then we have

hQ|P i
= hD|Uq† Up |Di.
hQ0 |P 0 i

The left-hand side of this equation has no |Di dependence. Thus we have
Uq† Up = cI for some complex number c. It follows that we can only have
hP 0 |Q0 i =
6 0 if Up and Uq are the same up to a global phase. However we
assumed that this is not the case and therefore hQ0 |P 0 i = 0. Hence

hQ|P i = 0.

This means the states |Qi and |P i are orthogonal.

Problem 15. Consider three two-dimensional Hilbert spaces H1 , H2 and

H3 . Consider the normalized product state
1 X
X 1 X
1
|ψi = cjk` |ji ⊗ |ki ⊗ |`i
j=0 k=0 `=0

in the product Hilbert space H1 ⊗ H2 ⊗ H3 . Let U1 , U2 , U3 be unitary

operators acting in these Hilbert spaces. By the First Fundamental Theo-
rem of invariant theory applied to U1 , U2 , U3 , any polynomial in cjk` which
is invariant under the action on |ψi of the local unitary transformation
U1 ⊗ U2 ⊗ U3 is a sum of homogeneous polynomials of even degree (say 2r).
For r = 1 we have
1 X
X 1 X
1
Pσ1 σ2 (c) = cj1 k1 `1 c∗j1 kσ `
1 (1) σ2 (1)
j1 =0 k1 =0 `1 =0

where σ1 and σ2 are permutations of 1. We denote by e the identity per-

mutation. For r = 2 we have
1 X
X 1 X
1 X
1 X
1 X
1
Pσ1 σ2 (c) = cj1 k1 `1 cj2 k2 `2 c∗j1 kσ ` c∗j2 kσ ` .
1 (1) σ2 (1) 1 (2) σ2 (2)
j1 =0 k1 =0 `1 =0 j2 =0 k2 =0 `2 =0

(i) Calculate the invariants.

(ii) Describe the connection with the partial traces

ρ1 := tr23 (|ψihψ|), ρ2 := tr31 (|ψihψ|), ρ3 := tr12 (|ψihψ|)

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 192

192 Problems and Solutions

of the density operator ρ := |ψihψ|.

Solution 15. (i) Obviously for the case r = 1 (degree 2) we only have
the identity permutation, i.e. σ1 = σ2 = e with e(1) = 1, e(2) = 2. Thus
we find only one invariant, namely
1 X
X 1 X
1
I0 = cjk` c∗jk` = hψ|ψi = 1
j=0 k=0 `=0

which is the normalization condition. For the case r = 2 (degree 4) we

find four linearly independent quartic invariants since e(1) = 1, e(2) = 2,
σ(1) = 2, σ(2) = 1. Thus
1 X
X 1 X
1 X
1 X
1 X
1
I1 = Pee (c) = cj1 k1 `1 c∗j1 k1 `1 cj2 k2 `2 c∗j2 k2 `2 = hψ|ψi2
j1 =0 k1 =0 `1 =0 j2 =0 k2 =0 `2 =0

1 X
X 1 X
1 X
1 X
1 X
1
I2 = Peσ (c) = cj1 k1 `1 c∗j1 k1 `2 cj2 k2 `2 c∗j2 k2 `1
j1 =0 k1 =0 `1 =0 j2 =0 k2 =0 `2 =0
1 X
X 1 X
1 X
1 X 1
1 X
I3 = Pσe (c) = cj1 k1 `1 c∗j1 k2 `1 cj2 k2 `2 c∗j2 k1 `2
j1 =0 k1 =0 `1 =0 j2 =0 k2 =0 `2 =0
1 X
X 1 X
1 X
1 X
1 X
1
I4 = Pσσ (c) = cj1 k1 `1 c∗j1 k2 `2 cj2 k2 `2 c∗j2 k1 `1 .
j1 =0 k1 =0 `1 =0 j2 =0 k2 =0 `2 =0

(ii) We find the invariants I2 = tr(ρ23 ), I3 = tr(ρ22 ), I4 = tr(ρ21 ).

Problem 16. Consider two Hilbert spaces Hreg and Hsys and the product
state
|ψi = (α|0reg i + β|1reg i) ⊗ |0sys i
in the Hilbert space Hreg ⊗ Hsys , where reg stands for register and sys for
system. Consider the swap operation (swap gate)
Uswap ((α|0reg i + β|1reg i) ⊗ |0sys i) = |0reg i ⊗ (α|0sys i + β|1sys i).
Discuss the operation on physical grounds.

Solution 16. Creating such a superposition could violate conservation

laws (for example charge) and in this case is forbidden by superselection
rules.

Problem 17. The Toffoli gate is the unitary operator acting as

UT |a, b, ci = |a, b, a · b + ci ≡ |ai ⊗ |bi ⊗ |b + ci
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 193

Unitary Transforms and Quantum Gates 193

in the Hilbert space C8 , where a, b, c ∈ { 0, 1 } and ab denotes the AND

operation of a and b. The addition + is modulo 2.
(i) Find the truth table.
(ii) Find the matrix representation for the standard basis.
The Toffoli gate is an extension of the CNOT gate.

Solution 17. (i) We have the truth table

a b c a b ab + c
0 0 0 0 0 0
0 0 1 0 0 1
0 1 0 0 1 0
0 1 1 0 1 1
1 0 0 1 0 0
1 0 1 1 0 1
1 1 0 1 1 1
1 1 1 1 1 0

(ii) The matrix representation of the Toffoli gate is given by the 8 × 8

permutation matrix  
0 1
I6 ⊕
1 0
where ⊕ denotes the direct sum.

Problem 18. The Fredkin gate is the unitary operator acting as

UF |c, x, yi = |c, cx + c̄y, c̄x + cyi

in the Hilbert space C8 , where c, x, y ∈ { 0, 1 }.

(i) Consider the cases c = 0 and c = 1.
(ii) Find the matrix representation for the standard basis.

Solution 18. (i) For c = 0 we have c̄ = 1. Therefore

cx = 0, c̄x = x, cy = 0, c̄y = y.

Thus
UF |0, x, yi = |0, y, xi.
For c = 1 we have c̄ = 0. Therefore cx = x, c̄x = 0, cy = y, c̄y = 0. Thus

UF |1, x, yi = |1, x, yi.

Consequently c is a control bit. If c = 0 then x and y swap around. If c = 1

then x and y stay the same.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 194

194 Problems and Solutions

(ii) The matrix representation of the Fredkin gate is given by the 8 × 8

permutation matrix
(1) ⊕ UN OT ⊕ I5
where ⊕ denotes the direct sum.

Problem 19. Consider the 8 × 8 matrix

eiα
U (α) = √ (I2 ⊗ I2 ⊗ I2 + iσ1 ⊗ σ1 ⊗ σ1 )
2
where α ∈ R.
(i) Show that U (α) is unitary.
(ii) Consider the standard basis |0i, |1i and the product state
     
0 0 0
|ψi = | ↓i ⊗ | ↓i ⊗ | ↓i ≡ ⊗ ⊗ .
1 1 1
Calculate the state U |ψi.
(iii) Consider U (α = 0) and the unitary 8 × 8 diagonal matrix
V = diag(ei3φ/2 , 1, 1, 1, 1, 1, 1, e−i3φ/2 ).
Calculate the state V U (α = 0)|ψi.
(iv) Calculate the state U (α = 0)V U (α = 0)|ψi.
(v) Let |ξ1 i = | ↓i ⊗ | ↓i ⊗ | ↓i, |ξ2 i = | ↑i ⊗ | ↑i ⊗ | ↑i. Calculate the
probabilities
|hξ1 |U (α = 0)V U (α = 0)|ψi|2 , |hξ2 |U (α = 0)V U (α = 0)|ψi|2 .

Solution 19. (i) Since σ1∗ = σ1 we have

e−iα
U ∗ (α) = √ (I2 ⊗ I2 ⊗ I2 − iσ1 ⊗ σ1 ⊗ σ1 ).
2
Since σ12 = I2 we obtain
1
U ∗ (α)U (α) = (I2 ⊗ I2 ⊗ I2 + I2 ⊗ I2 ⊗ I2 ) = I2 ⊗ I2 ⊗ I2 = I8 .
2
Thus U is a unitary matrix.
(ii) We find
eiα
U (α)|ψi = √ (I2 ⊗ I2 ⊗ I2 + iσ1 ⊗ σ1 ⊗ σ1 )(| ↓i ⊗ | ↓i ⊗ | ↓i)
2
eiα
= √ (| ↓i ⊗ | ↓i ⊗ | ↓i + i| ↑i ⊗ | ↑i ⊗ | ↑i)
2
eiα
= √ (| ↓i ⊗ | ↓i ⊗ | ↓i + eiπ/2 | ↑i ⊗ | ↑i ⊗ | ↑i).
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 195

Unitary Transforms and Quantum Gates 195

This is the GHZ state.

(iii) We find

1
V U (α = 0)|ψi = √ (e−i3φ/2 | ↓i ⊗ | ↓i ⊗ | ↓i + iei3φ/2 | ↑i ⊗ | ↑i ⊗ | ↑i)
2
1
= √ (cos(3φ/2)(| ↓i ⊗ | ↓i ⊗ | ↓i + i| ↑i ⊗ | ↑i ⊗ | ↑i)
2
−i sin(3φ/2)(| ↓i ⊗ | ↓i ⊗ | ↓i − i| ↑i ⊗ | ↑i ⊗ | ↑i)).

(iv) Using the result from (iii) we find

3 3
U (α = 0)V U (α = 0)|ψi = i cos( φ)| ↑i⊗| ↑i⊗| ↑i−i sin( φ)| ↓i⊗| ↓i⊗| ↓i.
2 2
(v) We obtain for the probabilities

1 1
|hξ1 |U (α = 0)V U (α = 0)|ψi|2 = sin2 (3φ/2) ≡ − cos(3φ)
2 2
and
1 1
|hξ2 |U (α = 0)V U (α = 0)|ψi|2 = cos2 (3φ/2) ≡ + cos(3φ).
2 2
Draw |hξ1 |U (α = 0)V U (α = 0)|ψi|2 and |hξ2 |U (α = 0)V U (α = 0)|ψi|2 as
functions of φ.

Problem 20. A quantum 2-torus is based on a C ∗ -algebra A generated

by the elements U1 and U2 with the relations

U1−1 = U1∗ , U2∗ = U2−1 , U1 U2 = zU2 U1

where z is a fixed complex unit. The algebra A will be commutative if

and only if z = 1, and in this case it describes the classical 2-torus. If z
is different from 1 then A describes a purely quantum object. This space
is called a quantum torus. A non commutative d-torus Tθd is a C ∗ -algebra
generated by d-unitaries U1 , U2 , . . . , Ud subject to the relations

Uα Uβ = e2πiθαβ Uβ Uα , α, β = 1, 2, . . . , d

where θ = (θαβ ) is a skew-symmetric matrix with real entries. Consider

the C ∗ -algebra given by the 2 × 2 matrices.
(i) Let    
0 1 0 1
U1 = , U2 = .
1 0 −1 0
Can we find z ∈ C such that U1 U2 = zU2 U1 ?
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 196

196 Problems and Solutions

(ii) Let    
1 1 1 0 1
U1 = √ , U2 = .
2 1 −1 1 0
Can we find z ∈ C such that U1 U2 = zU2 U1 ?

Solution 20. (i) Since

   
−1 0 1 0
U1 U2 = , U2 U1 =
0 1 0 −1

we have U1 U2 = eiπ U2 U1 .
(ii) Since
   
1 1 1 1 1 −1
U1 U2 = √ , U2 U1 = √
2 −1 1 2 1 1

we cannot find z ∈ C such that U1 U2 = zU2 U1 , but we find the unitary

matrix  
0 1
Z= = iσ2
−1 0
such that U1 U2 = ZU2 U1 .

Problem 21. Let α ∈ R. Find the unitary matrix

U (α) = exp(−iα(σ1 ⊗ σ1 + σ2 ⊗ σ2 + σ3 ⊗ σ3 )).

Solution 21. Since

[σ1 ⊗ σ1 , σ2 ⊗ σ2 ] = 04 , [σ1 ⊗ σ1 , σ3 ⊗ σ3 ] = 04 , [σ2 ⊗ σ2 , σ3 ⊗ σ3 ] = 04

we can write

U (α) = exp(−iα(σ1 ⊗ σ1 )) exp(−iα(σ2 ⊗ σ2 )) exp(−iα(σ3 ⊗ σ3 )).

It follows that

U (α) = (cos3 (α)−i sin3 (α))I4 −i sin(α) cos(α)eiα (σ1 ⊗σ1 +σ2 ⊗σ2 +σ3 ⊗σ3 ).

Problem 22. The Hadamard gate is defined by

1 1
UH |0i = √ (|0i + |1i), UH |1i = √ (|0i − |1i)
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 197

Unitary Transforms and Quantum Gates 197

and the CNOT gate is defined as

UCN OT (|ai ⊗ |bi) = |ai ⊗ |a ⊕ bi
where a, b ∈ { 0, 1 } and ⊕ is the XOR operation. Calculate the state
UCN OT (UH ⊗ I)(|0i ⊗ |0i), UCN OT (UH ⊗ I)(|0i ⊗ |1i)
UCN OT (UH ⊗ I)(|1i ⊗ |0i), UCN OT (UH ⊗ I)(|1i ⊗ |1i)
and discuss.

Solution 22. We obtain

1
UCN OT (UH ⊗ I)(|0i ⊗ |0i) = √ (|0i ⊗ |0i + |1i ⊗ |1i)
2
1
UCN OT (UH ⊗ I)(|0i ⊗ |1i) = √ (|0i ⊗ |1i + |1i ⊗ |0i)
2
1
UCN OT (UH ⊗ I)(|1i ⊗ |0i) = √ (|0i ⊗ |0i − |1i ⊗ |1i)
2
1
UCN OT (UH ⊗ I)(|1i ⊗ |1i) = √ (|0i ⊗ |1i − |1i ⊗ |0i).
2
These are the Bell states. Thus we generated the Bell states (which are
maximally entangled) from non-entangled states.

Problem 23. Consider the unitary 2 × 2 matrix

 
1 1 −1
U= √ .
2 1 1
Calculate the logarithm of U , i.e. log(U ) using
Z 1
log(U ) = (U − I2 )(t(U − I2 ) + I2 )−1 dt
0

to find B given by U = exp(B). This equation can be applied if the matrix

U has no eigenvalues on R− (the closed negative real axis). Set B = iK.
Find K.

Solution 23. The eigenvalues of U are given by

1 1
λ1 = √ (1 + i), λ2 = √ (1 − i).
2 2
Thus the condition to apply the equation is satisfied. We consider
√ first
the general case √
U = (ujk ) and then simplify to u11 = u22 = 1/ 2 and
u21 = −u12 = 1/ 2. We obtain
 
1 + t(u11 − 1) tu12
t(U − I2 ) + I2 =
tu21 1 + t(u22 − 1)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 198

198 Problems and Solutions

and
det(t(U − I2 ) + I2 ) = d(t) = 1 + t(−2 + tr(U )) + t2 (1 − tr(U ) + det(U )).
Let X ≡ det(U ) − tr(U ) + 1. Then
 
−1 1
tX + u11 − 1 u12
(U − I2 )(t(U − I2 ) + I2 ) = .
u21 d(t)tX + u22 − 1
√ √
With u11 = u22 = 1/ 2, u21 = −u12 = 1/ 2 we obtain
√ √
d(t) = 1 + t(−2 + 2) + t2 (2 − 2)
√
and X ≡ det(U ) − tr(U ) + 1 = 2 − 2. Thus the matrix takes the form
 √ √ √ 
1 t(2 − 2) + √1/ 2 − 1 √−1/ 2 √ .
d(t) 1/ 2 t(2 − 2) + 1/ 2 − 1
Since
 √ √ !1
Z 1
1 2  2(2 − 2)t + 2 − 2  √ π
dt = √ arctan √  = 2
0 d(t) 2 2  4
0

and Z 1
t 1 π
dt = √
0 d(t) 24
we obtain
   
0 −π/4 0 iπ/4
B= , K = −iB = .
π/4 0 −iπ/4 0

Problem 24. Let { |0i, |1i, . . . , |n − 1i } be an orthonormal basis in the

Hilbert space Cn .
(i) Is the linear operator
n−1
X
Zn := exp(2πij/n)|jihj|
j=0

unitary?
(ii) Can the operator Zn be expressed as the exponent of a hermitian op-
erator?

Solution 24. (i) Since

n−1
X
Zn∗ = exp(−2πij/n)|jihj|
j=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 199

Unitary Transforms and Quantum Gates 199

we find Zn Zn∗ = I. Therefore Zn is unitary.

(ii) Since Zn is unitary we can find a hermitian operator, say K̂, such that
Zn = exp(iK̂). We obviously find
n−1
X
Zn = exp(2πiθ̂/n), θ̂ := j|jihj|.
j=0

The operator θ̂ is the SU (2) phase operator.

Problem 25. The four Bell states are given by

1 1
|φ+ i = √ (|0i ⊗ |0i + |1i ⊗ |1i), |φ− i = √ (|0i ⊗ |0i − |1i ⊗ |1i)
2 2
1 1
|ψ + i = √ (|0i ⊗ |1i + |1i ⊗ |0i), |ψ − i = √ (|0i ⊗ |1i − |1i ⊗ |0i).
2 2
Show that the Bell states can be transformed to each other under local
unitary transformations (i.e. the Kronecker product U ⊗ V of two unitary
2 × 2 matrices U and V ). Hint. Consider the Pauli spin matrices and the
2 × 2 identity matrix.

Solution 25. We find

|φ− i = (I2 ⊗ σ3 )|φ+ i = (σ3 ⊗ I2 )|φ+ i

|ψ + i = (I2 ⊗ σ1 )|φ+ i = (σ1 ⊗ I2 )|φ+ i
|ψ − i = (I2 ⊗ (−iσ2 ))|φ+ i = (iσ2 ⊗ I2 )|φ+ i.

Problem 26. Consider the unitary matrix

 
0 0 1
V = 0 1 0.
1 0 0

Find the hermitian matrix K such that V = exp(iK).

Solution 26. The matrix V is also hermitian. We calculate the eigenval-

ues and normalized eigenvectors of V . From the normalized eigenvectors
we construct a unitary matrix W such that W ∗ V W is a diagonal matrix.
Then ∗
W ∗ V W = W ∗ eiK W = eiW KW = eiL
where L = W ∗ KW . Since W ∗ V W is a diagonal matrix L is also a diagonal
matrix. Thus (W ∗ V W )jj = ei`jj . Finally we find K from K = W LW ∗ .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 200

200 Problems and Solutions

Now the eigenvalues of V are given by λ1 = 1, λ2 = 1, λ3 = −1 with the

corresponding normalized eigenvectors
     
0 1 1
1 1
1, √ 0, √  0 .
0 2 1 2 −1

This leads to the unitary (orthogonal) matrix

 √ √ 
0 1/ 2 1/ 2
W = 1 0√ 0√  .
0 1/ 2 −1/ 2

Thus we obtain the diagonal matrix

 
1 0 0
W ∗V W =  0 1 0  = eiL .
0 0 −1

From W ∗ V W = exp(iL) we find the equations 1 = ei`11 , 1 = ei`22 , −1 =

ei`33 . The solution is `11 = `22 = 0, `33 = π. Hence we obtain the hermitian
matrix  
1 0 −1
π
K = W LW ∗ =  0 0 0  .
2
−1 0 1

Problem 27. Let U be an n×n unitary matrix. Show that if the bipartite
states |ψi, |φi ∈ Cn ⊗ Cm satisfy |φi = (U ⊗ Im )|ψi, then the ranks of the
corresponding reduced density matrices satisfy

r(ρψ φ
1 ) ≥ r(ρ1 ), r(ρψ φ
2 ) ≥ r(ρ2 ).

Solution 27. We consider the Schmidt decomposition of the state |ψi

s q
X
|ψi = λψ
j |ji ⊗ |ji, λψ
j > 0, s ≤ min(n, m)
j=1

where s is the number of non vanishing terms in the Schmidt decomposition.

We write the unitary operator as
n
X
U= |µj ihj|
j=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 201

Unitary Transforms and Quantum Gates 201

where |µj i ∈ Cn . Then we find that

s
X s
X
ρψ
1 = |jihj|, ρφ1 = U ρψ ∗
1U = |µj ihµj |.
j=1 j=1

Thus r(ρφ1 ) ≤ s. The second inequality follows from the fact that for any
bipartite state r(ρ1 ) = r(ρ2 ).

Problem 28. Consider the 4 × 4 hermitian matrix

R := σ1 ⊗ σ1 + σ2 ⊗ σ2 + σ3 ⊗ σ3
and the Hamilton operator
1
Ĥ(t) = ~ω(t)R.
2
Let Z T
1
φ= ω(t)dt
2 0
and
! ! !
Z T Z T
1
U (T ) = exp −i Ĥ(t)dt/~ = exp − ω(t)dt R
0 2 0

= exp(−iφR).
Calculate U (T ) and express it using the swap gate
1 0 0 0
 
0 0 1 0
Usw =  .
0 1 0 0
0 0 0 1

Solution 28. Note that Usw = I4 + R. We have

1 1
U (T ) = e−iφR = e−iφ (3I4 + R) + e−3iφ (I4 − R)
 4 4 
iφ i
=e cos(2φ)I4 − sin(2φ)(I4 + R)
2
= eiφ (cos(2φ)I4 − i sin(2φ)Usw ).

Problem 29. Consider the Bell basis in the form

1 0 0 −i
      
1  0  1  1  1  −i  1  0 
√  , √  , √  , √  .
2 0 2 −i 2 1 2 0
−i 0 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 202

202 Problems and Solutions

From the Bell states we form the matrix

1 0 0 −i
 
1  0 1 −i 0 
B=√  .
2 0 −i 1 0
−i 0 0 1

(i) Show that B is unitary. Calculate B 2 , B 3 , B 4 , . . . , B 8 and show that B k

(k = 1, 2 . . . , 8) form a group under matrix multiplication.
(ii) Let γ1 and γ4 the Dirac gamma matrices

0 0 0 −i 1 0 0 0
   
0 0 −i 0  0 1 0 0 
γ1 =  , γ4 =  .

0 i 0 0 0 0 −1 0
i 0 0 0 0 0 0 −1

Calculate exp(πγ4 γ1 /4) and show that exp(πγ4 γ1 /4) = B.

(iii) Let T2 = B and

T1 = exp(−iπI2 ⊗ σ3 /4), T3 = exp(−iπσ3 ⊗ I2 /4).

Calculate T1 and T3 . Show that T1 T2 T1 = T2 T1 T2 , T3 T2 T3 = T2 T3 T2 and

T1 T3 = T3 T1 , i.e. we have a braid like relation.

Solution 29. (i) We have B ∗ B = I4 . Thus B ∗ = B −1 and B is unitary.

We define
0 0 0 1
 
0 0 1 0
N :=  .
0 1 0 0
1 0 0 0

Thus N 2 = I4 and we can write B = √1 (I4 − iN ). We find

2

1
B 2 = (I4 − iN )(I4 − iN ) = −iN
2
3 1 1
B = √ (I4 − iN )(−iN ) = − √ (I4 + iN )
2 2
4 2 2
B = B B = (−iN )(−iN ) = −I4
B 5 = −B
B 6 = −B 2 = iN
1
B 7 = −B 3 = √ (I4 + iN )
2
B 8 = B 4 B 4 = I4 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 203

Unitary Transforms and Quantum Gates 203

(ii) We have

0 0 0 −i
 
 0 0 −i 0  2
γ4 γ1 =   = −iN = B .
0 −i 0 0
−i 0 0 0

Thus γ4 γ1 = −iN and (γ4 γ1 )2 = −I4 , (γ4 γ1 )3 = iN , (γ4 γ1 )4 = I4 . Using

this result we find
1
exp(πγ4 γ1 /4) = I4 cos(π/4) − iN sin(π/4) = √ (I4 − iN ) = B.
2
(iii) Using that exp(−iπ/2) = −i we find

−i 0 0 −i
 
1  0 −i −i 0 
T3 T2 T3 = √  .
2 0 −i i 0
−i 0 0 i
√ √
Using that e−iπ/4 − eiπ/4 = −i 2 and e−iπ/4 + eiπ/4 = 2 we find
−i 0 0 −i
 
1  0 −i −i 0 
T2 T3 T2 = √  .
2 0 −i i 0
−i 0 0 i
Thus T2 T3 T2 = T3 T2 T3 . Analogously we find T1 T2 T1 = T2 T1 T2 . Obviously
T1 T3 = T3 T1 since T1 and T3 are diagonal matrices.

Problem 30. Consider the swap gate

1 0 0 0
 
0 0 1 0
Uswap =  .
0 1 0 0
0 0 0 1
Can the swap gate be written as
1 0 0 0
 
0 1 0 0  −1
Uswap = R R
0 0 −1 0
0 0 0 1
where R is an orthogonal matrix?

Solution 30. The matrix Uswap is not only unitary but also hermitian
with eigenvalues +1, +1, +1, −1. To construct the matrix R we just find
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 204

204 Problems and Solutions

the normalized eigenvectors of U . This yields

1 0√ 0√ 0
 
−1 ∗  0 1/√2 1/ √2 0
R=R =R = .
0 1/ 2 −1/ 2 0
0 0 0 1

Problem 31. Let d ≥ 2 and |0i, |1i, . . . , |d − 1i be an orthonormal basis

in the Hilbert space Cd . Let |ψi, |φi be two normalized states in Cd . Let
d−1
X
S= ((|jihk|) ⊗ (|kihj|)).
j,k=0

Show that S(|ψi ⊗ |φi) = |φi ⊗ |ψi. Thus S is a swap operator.

Solution 31. We have

d−1
X
S(|ψi ⊗ |φi) = ((|jihk|) ⊗ (|kihj|))(|ψi ⊗ |φi)
j,k=0
d−1
X d−1
X
= hk|ψihj|φi(|ji ⊗ |ki) = hj|ψihk|φi(|ki ⊗ |ji)
j,k=0 j,k=0
= |φi ⊗ |ψi.

Problem 32. Let σ1 , σ2 , σ3 be the Pauli spin matrices and

σj,1 = I2 ⊗ · · · ⊗ I2 ⊗ σ1 ⊗ I2 ⊗ · · · ⊗ I2

where σ1 is a the j position (counting from left to right) and j = 1, 2, . . . , N .

Analogously we have σj,2 and σj,3 . Let

σj,± := σj,1 ± iσj,2 .

Consider the unitary matrix

 
n
X
U := exp i χj σj,+ σj,−  .
j=1

Find U σj,1 U ∗ , U σj,− U ∗ , U σj,3 U ∗ .

Solution 32. We obtain

U σj,1 U ∗ = eiχj σj,+ , U σj,− U ∗ = e−iχj σj,− , U σj,3 U ∗ = σj,3 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 205

Unitary Transforms and Quantum Gates 205

Problem 33. Let σ1 , σ3 be the Pauli spin matrices and

1
Π= (I2 + σ3 ) ⊗ I2 .
2
Show that Π is a projection matrix. Show that the 4 × 4 matrix
1 1
U= (I2 + σ3 ) ⊗ I2 + (I2 − σ3 ) ⊗ σ1
2 2
is unitary.

Solution 33. We have Π = Π∗ and Π2 = Π and

 
0 1
U = I2 ⊕ .
1 0

Thus U = U ∗ , U 2 = I4 and U is unitary.

Problem 34. The vectors

    

1 0 1
1 1
v1 = √  0  , v2 =  1  , v3 = √  0 
2 1 0 2 −1

form an orthonormal basis in the Hilbert space C3 . Find the unitary ma-
trices U12 , U23 , U31 such that U12 v1 = v2 , U23 v2 = v3 , U31 v3 = v1 . Then
calculate U31 U23 U12 and the matrix V = λ1 v1 v1∗ + λ2 v2 v2∗ + λ3 v3 v3∗ , where
the complex numbers λ1 , λ2 , λ3 satisfy λ1 λ1 = 1, λ2 λ2 = 1, λ3 λ3 = 1.
Show that the matrix is unitary.

Solution 34. We obtain

   √ √ 
0√ 1 0√ 0 1/ 2 1/ 2
U12 =  1/√2 0 1/ √2  , U23 =  1 0√ 0√  ,
1/ 2 0 −1/ 2 0 −1/ 2 1/ 2

U31 = I2 ⊕ (−1).
Obviously U31 U23 U12 = I3 . Now
     
1 0 1
λ1   λ3 
V = 0 ( 1 0 1 ) + λ2  1  ( 0 1 0)+ 0 (1 0 −1 ) .
2 2
1 0 −1

It follows that  
λ + λ3 0 λ1 − λ3
1 1
V = 0 2λ2 0 
2
λ1 − λ3 0 λ1 + λ3
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 206

206 Problems and Solutions

and V V ∗ = I3 , i.e. the matrix V is unitary.

Problem 35. Let φ ∈ R. Consider the 4 × 4 matrix

0 1 0 0
 
 0 0 1 0
A(φ) =  .
0 0 0 1
eiφ 0 0 0
(i) Is the matrix unitary?
(ii) Find the eigenvalues and normalized eigenvectors of A(φ).
(iii) Can the matrix be written as the Kronecker product of two 2 × 2
matrices?

Solution 35. (i) Yes. We have A(φ)A∗ (φ) = I4 .

(ii) The eigenvalues are ieiφ/4 , −eiφ/4 , −ieiφ/4 , eiφ/4 with the corresponding
normalized eigenvectors
1 1 1 1
       
1  ie iφ/4 iφ/4 iφ/4 iφ/4
 1  −e  1  −ie  1e
iφ/2  ,  iφ/2  , iφ/2  ,  iφ/2  .

2 −e 2 e 2 −e 2 e
 
−ie3iφ/4 −e3iφ/4 ie3iφ/4 e3iφ/4
(iii) No.

Problem 36. (i) Let A be an n × n matrix over C and Π be an m × m

projection matrix. Let z ∈ C. Calculate exp(z(A ⊗ Π)).
(ii) Let A1 , A2 be n × n matrices over C. Let Π1 , Π2 be m × m projection
matrices with Π1 Π2 = 0. Calculate exp(z(A1 ⊗ Π1 + A2 ⊗ Π2 )).
(iii) Use the result from (ii) to find the unitary matrix U (t) = exp(−iĤt/~),
where Ĥ = ~ω(A1 ⊗ Π1 + A2 ⊗ Π2 ) and we assume that A1 and A2 are
hermitian matrices.
(iv) Apply the result of (iii) to
   
1 1 1 1 1 −1
A1 = σ 1 , Π1 = , A2 = σ3 , Π2 = .
2 1 1 2 −1 1

Solution 36. (i) We find exp(z(A ⊗ Π)) = In ⊗ Im + (ezA − In ) ⊗ Π.

(ii) We obtain
exp(z(A1 ⊗ Π1 + A2 ⊗ Π2 )) = In ⊗ Im + (ezA1 − In ) ⊗ Π1 + (ezA2 − In ) ⊗ Π2 .
(iii) Since U (t) = exp(−iωt(A1 ⊗ Π1 + A2 ⊗ Π2 )) with z = −iωt we obtain
the unitary matrix
U (t) = In ⊗ Im + (e−iωtA1 − In ) ⊗ Π1 + (e−iωtA2 − In ) ⊗ Π2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 207

Unitary Transforms and Quantum Gates 207

(iv) With m = n = 2, A1 = σ1 and A2 = σ3 we find the unitary matrix

U (t) = I2 ⊗ I2 + (e−iωtσ1 − I2 ) ⊗ Π1 + (e−iωtσ3 − I2 ) ⊗ Π2 .

With exp(zσ1 ) = cosh(z)I2 + sinh(z)σ1 , exp(zσ3 ) = cosh(z)I2 + sinh(z)σ3

and z = −iωt and sinh(−iωt) ≡ −i sin(ωt), cosh(−iωt) ≡ cos(ωt) we obtain

U (t) = I2 ⊗ I2 + (I2 cos(ωt) − iσ1 sin(ωt) − I2 ) ⊗ Π1

+(I2 cos(ωt) − iσ3 sin(ωt) − I2 ) ⊗ Π2 .

Programming Problems

Problem 1. Given the standard basis e1 , e2 , e3 , e4 in C4 . Consider the

unitary matrices
1 0 0 0
 
0 1 0 0
UCN OT =   , VH = UH ⊗ I2
0 0 0 1
0 0 1 0
where UH is the Hadamard gate
 
1 1 1
UH =√ .
2 1 −1

Show that UCN OT VH ej , (j = 1, 2, 3, 4) will provide the Bell states.

Solution 1. The Maxima program will do the job

/* BellUnitary.mac */
e1: matrix([1],[0],[0],[0]); e2: matrix([0],[1],[0],[0]);
e3: matrix([0],[0],[1],[0]); e4: matrix([0],[0],[0],[1]);
UCNOT: matrix([1,0,0,0],[0,1,0,0],[0,0,0,1],[0,0,1,0]);
UH: matrix([1/sqrt(2),1/sqrt(2)],[1/sqrt(2),-1/sqrt(2)]);
I2: matrix([1,0],[0,1]);
KUHI2: kronecker_product(UH,I2);
b1: UCNOT . KUHI2 . e1; b2: UCNOT . KUHI2 . e2;
b3: UCNOT . KUHI2 . e3; b4: UCNOT . KUHI2 . e4;

The output is
1 0 0 1
       
1 0 1 1 1  1  1  0 
√  , √  , √  , √  .
2 0 2 1 2 −1 2 0
1 0 0 −1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 208

208 Problems and Solutions

Problem 2. Consider the normalized vector in C4

1 T
v= (i −i i −i ) .
2
Show that U = I4 − 2vv∗ is a unitary matrix. Find the eigenvalues of U .

Solution 2. The following Maxima program will do the job.

/* unitary.mac */
v: (1/2)*matrix([%i],[-%i],[%i],[-%i]);
vT: transpose(v); vTC: conjugate(vT);
I4: matrix([1,0,0,0],[0,1,0,0],[0,0,1,0],[0,0,0,1]);
U: I4-2*(v . vTC); UT: transpose(U); UTC: conjugate(UT);
R: U . UTC;
eigenvalues(U);

The eigenvalues are −1 (1 times) and +1 (3 times).

7.3 Supplementary Problems

Problem 1. (i) Let φ ∈ R. Show that the matrix

0 e−iφ
 
A(φ) =
eiφ 0

hermitian and unitary. Find the eigenvalues and normalized eigenvectors

of A(φ). Let I2 be the 2 × 2 unit matrix. Find the eigenvalues of A(φ) ⊗ I2 .

Problem 2. Let U be a unitary and hermitian n × n matrix. Show that

1 1
Π+ = (In + U ), Π− = (In − U )
2 2
are projection matrices. Show that Π+ Π− = 0n .

Problem 3. Find all the square roots of the 2 × 2 identity matrix. For
example the Pauli spin matrices σ1 , σ2 , σ3 are solutions.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 209

Chapter 8

Entropy

8.1 Introduction
For any density operator (where hψj |ψk i = δjk )
n
X n
X
ρ= λj |ψj ihψj |, λj ≥ 0, λj = 1
j=1 j=1

the von Neumann entropy is defined as

S(ρ) := −tr(ρ log(ρ))

or equivalently
n
X
S(ρ) := − λj log(λj )
j=1

with
0 log(0) = 0, 1 log(1) = 0.
Thus the von Neumann entropy is equal to the Shannon entropy of the
eigenvalues. We have S(ρ) ≥ 0 with equality iff ρ is a pure state, i.e.
ρ = |ψihψ|. Furthermore we have the inequality

S(ρ) ≤ log(n)

209
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 210

210 Problems and Solutions

where n is the dimension of ρ. We find equality iff

1
ρ= I
n
where I is the identity operator. The entropy is unchanged under unitary
transformation
S(U ρU ∗ ) = S(ρ).
For a joint system AB we have
S(ρAB ) ≤ S(ρA ) + S(ρB ).
For density operators ρ and σ we have the quantum relative entropy
S(ρ||σ) := tr(ρ log(ρ)) − tr(ρ log(σ)).
We have Klein’s inequality
S(ρ||σ) ≥ 0
with equality iff ρ = σ.

Let A, B be n × n hermitian matrices acting in the Hilbert space Cn .

Assume that the eigenvalues of A are pairwise different and analogously
for B. Then the normalized eigenvectors |αj i (j = 1, . . . , n) of A form an
orthonormal basis in Cn and analogously for B the normalized eigenvectors
|βj i (j = 1, . . . , n) form an orthonormal basis in Cn . Let |ψi be a normalized
state in Cn . Then there are n possible outcomes for measurements of each
observable and the probabilities pj (A, |ψi), pj (B, |ψi) (j = 1, . . . , n) are
given by
pj (A, |ψi) := |hψ|αj i|2 , pj (B, |ψi) := |hψ|βj i|2 .
Let H|ψi (X) be the Shannon information entropy
n
X
H|ψi (X) := − pj (X, |ψi) ln(pj (X, |ψi))
j=1

corresponding to the probability distribution {pj (X, |ψi)} (j = 1, . . . , n).

The (Maassen-Uffink) entropic uncertainty relation is given by
H|ψi (A) + H|ψi (B) ≥ −2 ln( max |hαj |βk i|) > 0.
1≤j,k≤n

Note that the right-hand side does not involve the state |ψi.

The (Landau-Pollak) uncertainty relation states that

p p
arccos( PA ) + arccos( PB ) ≥ arccos( max |hαj |βk i|)
1≤j,k≤n

where
PA := max pj (A, |ψi), PB := max pj (B, |ψi).
1≤j≤n 1≤j≤n
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 211

Entropy 211

8.2 Solved Problems

Problem 1. (i) Consider the density matrix (pure state)
   
1 1 1 1 1 1
ρ= ≡√ √ (1 1).
2 1 1 2 1 2
Find the von Neumann entropy S(ρ) = −tr(ρ log2 (ρ)).
(ii) Consider the density matrix (mixed state)
 
1/2 0
ρ= .
0 1/2

Find the von Neumann entropy S(ρ) = −tr(ρ log2 (ρ)).

Solution 1. (i) The eigenvalues of ρ are 0 and 1. Hence S(ρ) = 0.

(ii) Since log2 (1/2) = −1 we obtain S(ρ) = 1. Note that ρ describes a
mixed state.

Problem 2. Let W be a positive semidefinite n × n matrix over C with

tr(W ) = 1. Let λ1 , λ2 ,. . . , λn be the eigenvalues (which obviously are real
and nonnegative) and w1 , w2 , . . . , wn be the corresponding normalized
eigenvectors of W . We can assume that the normalized eigenvectors form
an orthonormal basis in Cn . If eigenvalues are degenerate and the cor-
responding normalized eigenvectors are not orthogonal we can apply the
Gram-Schmidt algorithm. Calculate tr(W ln(W )).

Solution 2. Let U be a unitary matrix such that U −1 W U is a diag-

onal matrix. Note that U −1 = U ∗ . Obviously the unitary matrix U is
constructed from the normalized eigenvectors of W . Then we have

tr(W ln W ) = tr(U −1 (W ln(W ))U ) = tr(U −1 W U U −1 (ln(W ))U )

= tr(U −1 W U ln(U −1 W U )).

The diagonal elements of U −1 W U are the nonnegative real eigenvalues of

W . Consequently
Xn
tr(W ln(W )) = λj ln(λj ).
j=1

Note that 0 · ln(0) = 0 if an eigenvalue of W is 0.

Problem 3. Let W be a positive semidefinite n × n matrix over C with

tr(W ) = 1. Let λ1 , λ2 , . . . , λn be the eigenvalues (which obviously are real
and nonnegative) and w1 , w2 , . . . , wn be the corresponding normalized
eigenvectors (column vectors) of W . We can assume that the normalized
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 212

212 Problems and Solutions

eigenvectors form an orthonormal basis in Cn . If eigenvalues are degenerate

and the corresponding normalized eigenvectors are not orthogonal we can
apply the Gram-Schmidt algorithm. Let W f be a positive definite n × n ma-
trix over C with tr(W
f ) = 1. Let λe1 , λ
e2 , . . . , λ
en be the eigenvalues (which
obviously are real and positive) and w e 1, w
e 2, . . . , w
e n be the corresponding
normalized eigenvectors of W . We can assume that the normalized eigen-
f
vectors form an orthonormal basis in Cn . If eigenvalues are degenerate
and the corresponding eigenvectors are not orthogonal we can apply the
Gram-Schmidt algorithm. Calculate

tr(W ln(W
f )).

What happens if we allow the matrix W

f to be positive semidefinite?

Solution 3. Let U e be a unitary matrix such that U e −1 W

fUe is a diag-
−1 ∗
onal matrix. Note that Ue = U . Obviously the unitary matrix U
e e is
constructed from the normalized eigenvectors of W . Then we have
f

tr(W ln(W e −1 (W ln(W

f )) = tr(U f ))U
e ) = tr(Ue −1 W UeUe −1 (ln W
f )U
e)
e −1 W U
= tr(U e −1 W
e ln(U fUe )) = tr(Ue −1 W Ue ln(WfD ))

where W e −1 W
fD = U fUe is a diagonal matrix. Using the spectral representa-
tion of W
X n
W = λj wj wj∗
j=1

we obtain
n
X
tr(W ln(W
f )) = e −1 (wj w∗ )U
λj tr(U e ln(W
fD )).
j
j=1

Since
n
X n
X
U
e= e ` e∗` ,
w e∗ =
U e k∗
ek w
`=1 k=1

where { ek : k = 1, 2, . . . , n } is the standard basis in Cn we obtain

n
X
tr(W ln(W
f )) = e −1 (wj w∗ )U
λj tr(U e ln(W
fD ))
j
j=1
Xn Xn X
n
= e k∗ (wj wj∗ )w
λj tr(ek w e ` e∗` ln(W
fD )).
j=1 k=1 `=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 213

Entropy 213

To calculate the trace we use the standard basis { ej : j = 1, 2, . . . , n }.

Thus
n X
X n X
n X
n
tr(W ln(W
f )) = λj (e∗r ek w
e k∗ (wj wj∗ )w
e ` e∗` ln(W
fD )er )
j=1 k=1 `=1 r=1
Xn X n X n X n
= e k∗ (wj wj∗ )w
λj (δrk w e ` e∗` ln(W
fD )er )
j=1 k=1 `=1 r=1
n X
X n X
n
= e k∗ (wj wj∗ )w
λ j (w e ` e∗` ln(W
fD )ek )
j=1 k=1 `=1
n X
X n X
n
= e k∗ (wj wj∗ )w
λ j (w e ` δ`k ln(λ
ek ))
j=1 k=1 `=1
Xn X n
= e k∗ (wj wj∗ )w
λ j (w e k ln(λ
ek )).
j=1 k=1

Consequently
n X
X n
tr(W ln(W
f )) = e k∗ wj )(wj∗ w
λj (w e k ) ln(λ
ek ).
j=1 k=1

Note that we k∗ wj = (wj∗ we k )∗ . Can the condition on W

f be extended to
positive semidefinite, i.e. some of the eigenvalues of W f could be zero?
Assume that one eigenvalue of W is 0, say λp = 0. Thus we should have
f e
n
X
e p∗ (wj wj∗ )w
λj w e p∗ W w
ep = w ep = 0
j=1

in order to apply 0 · ln 0 = 0. This is in general not true. Consider, for

example the density matrices
   
1/2 0 f= 1 0 .
W = , W
0 1/2 0 0

Problem 4. The relative entropy of entanglement for bipartite states (say,

A and B), where the quantum state is described by the density matrix W ,
is defined as
Er (W ) := min S(W kW f)
f ∈D
W
where D is the set of all convex combinations of separable density matrices
f = WA ⊗ WB ) and
(i.e. W

S(W kW
f ) := tr(W (log2 (W ) − log2 (W
f ))) ≡ tr(W log2 (W ) − W log2 (W
f ))
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 214

214 Problems and Solutions

is the quantum relative entropy. Let W

fmin denote the separable state that
minimizes the relative entropy. Thus to calculate Er (W ) is to find the state
W
fmin . Consider the Bell state

1
|Φ+ i = √ (|0i ⊗ |0i + |1i ⊗ |1i)
2

and thus W = |Φ+ ihΦ+ |. Let

f = 1 (|0ih0| + |1ih1|) ⊗ 1 (|0ih0| + |1ih1|).

W
2 2

Calculate S(W kW
f ).

Solution 4. Using the standard basis in C4 we obtain the density matrices

1 0 0 1
 
1 0 0 0 0
W = |Φ+ ihΦ+ | = 

2 0 0 0 0

1 0 0 1

and
f = 1 I4
W
4
where I4 is the 4 × 4 identity matrix. To calculate tr(W log2 (W )) and
tr(W log2 (W
f )) we need the eigenvalues and normalized eigenvectors of W
and W . For the density matrix W we find the eigenvalues λ1 = 1, λ2 = 0,
f
λ3 = 0, λ4 = 0 with the corresponding normalized eigenvectors

1 1 0 0
      
1 0 1  0  1 0
w1 = √   , w2 = √  , w3 =   , w4 =   .
2 0 2 0 0 1
1 −1 0 0

Thus we find
4
X
tr(W log2 (W )) = λj log2 (λj ) = 0
j=1

where we used 0 log2 (0) = 0 and 1 log2 (1) = 0. For W f we obtain the
eigenvalues λ1 = 1/4, λ2 = 1/4, λ3 = 1/4, λ4 = 1/4 and the corresponding
e e e e
normalized eigenvectors (standard basis)

1 0 0 0
       
0 1 0 0
w
e1 =  , w
e2 =  , w
e3 =  , w
e4 =  .
0 0 1 0
0 0 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 215

Entropy 215

Since
4 X
X 4
tr(W log2 (W
f )) = e k∗ wj )(wj∗ w
λj (w e k ) log2 (λ
ek )
j=1 k=1

and
1 1
w1∗ w
e1 = √ , w1∗ w
e 2 = 0, w1∗ w
e 3 = 0, w1∗ w
e4 = √
2 2
1 1
w2∗ w
e1 = √ , w2∗ w
e 2 = 0, w2∗ w
e 3 = 0, w2∗ w
e4 = −√
2 2
w3∗ w
e 1 = 0, w3∗ w
e 2 = 1, w3∗ w
e 3 = 0, w3∗ w
e4 = 0
w4∗ w
e 1 = 0, w4∗ w
e 2 = 0, w4∗ w
e 3 = 1, w4∗ w
e4 = 0
f )) = −2. Consequently, S(W ||W
we obtain tr(W log2 (W f ) = 2.

Problem 5. Let ρAB be a density matrix defined on a (N ×N )-dimensional

Hilbert space H ⊗ H. The classical information capacity is defined as

C(ρ) := log2 (N ) + S(ρB ) − S(ρAB )

where ρB is the reduced quantum state obtained by ρB = trA (ρAB ) and

S(ρ) is the von Neumann entropy of a quantum state (density matrix)
S(ρ) = −tr(ρ log2 (ρ)). Consider the Bell state

1
|ψi = √ (|0i ⊗ |0i + |1i ⊗ |1i).
2
Calculate the density matrix ρAB , the reduced density matrix ρB and then
the classical information capacity.

Solution 5. We find the density matrix (pure state)

1 0 0 1
 
1 0 0 0 0
ρAB = |ψihψ| =  .
2 0 0 0 0
1 0 0 1

Thus we obtain the density matrix (mixed state)

 
1/2 0
ρB = .
0 1/2

With N = 2 and S(ρAB ) = 0 it follows that

C(ρ) = log2 (N ) + S(ρB ) − S(ρAB ) = 2.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 216

216 Problems and Solutions

Problem 6. The quantum relative entropy between two density operators

ρ and σ is defined by

Sb (ρ||σ) := tr(ρ logb (ρ) − ρ logb (σ)) = −Sb (ρ) − tr(ρ logb (σ)).

Here Sb is the von Neumann entropy where the log is taken with the base
b. Show that Sb (ρ||σ) ≥ 0. This inequality is known as Klein’s inequality.

Solution 6. Consider the term ρ logb (σ) where ρ and σ are density oper-
ators on a finite-dimensional Hilbert space of dimension n. Let λj and |φj i
be the corresponding eigenvalues and (orthonormal) eigenstates of ρ. Sim-
ilarly, let µk and |ψk i be the corresponding eigenvalues and (orthonormal)
eigenstates of σ. Thus we have
  !
Xn n
X
ρ log2 (σ) =  λj |φj ihφj | logb µk |ψk ihψk |
j=1 k=1
n
X
= λj logb (µk )hφj |ψk i|φj ihψk |.
j,k=1

Taking the trace using the basis {|ψk i, k = 1, 2, . . . , n} yields

n
X
tr(ρ logb (σ)) = λj (logb (µk ))hφj |ψk ihψl |φj ihψk |ψl i
j,k,l=1
Xn
= λj (logb (µk ))hφj |ψk ihψk |φj i
j,k=1
Xn
= λj (logb (µk ))|hφj |ψk i|2 .
j,k=1

Thus we obtain
n
X n
X
Sb (ρ||σ) = λj logb (λj ) − λj (logb (µk ))|hφj |ψk i|2
j=1 j,k=1
n n n
!
|hφ |ψ i|2
X X Y
= λj logb (λj ) − λj logb µk j k
j=1 j=1 k=1
Xn Xn
= λj logb (λj ) − λj logb (νj )
j=1 j=1

where !
n
|hφ |ψ i|2
Y
νj := µk j k .
k=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 217

Entropy 217

If νj = 0 for some j where λj 6= 0 then SB (ρ||σ) ≥ 0 trivially. We assume

νj = 0 if and only if λj = 0. Since
n
X n
X n
X
µk = |hφj |ψk i|2 = |hφj |ψk i|2 = 1
k=1 k=1 j=1

we find that
n
X
0 ≤ νj ≤ 1, α := νj ≤ 1.
j=1

We assume that the λj are ordered in non-decreasing order and m ≤ n is

chosen such that λj = 0 iff j > m. We determine the maximum value of
m
X m
X
λj logb (xj ), xj = α
j=1 j=1

which can be formulated as a Lagrange multiplier problem. Thus we find

the critical points of the function
 
m
X Xm
f (x1 , . . . , xm ) := λj logb (xj ) − θ  xj − α 
j=1 j=1

where θ is the Lagrange multiplier. We obtain

∂f λj
= − θ.
∂xj xj ln(b)

Since ∂f /∂xj = ∂f /∂xk = 0 for the critical points we have, since λj 6= 0

and λk 6= 0, xk = λk xj /λj . Inserting xk 6= 0 into the constraint yields
n n
X X λj
xj = xk + xk = α.
j=1
λ
j=1 k
j6=k

Pn Pm
Thus xk = αλk , since j=1 λj = 1. Since j=1 λj logb (xj ) is unbounded
from below we have a maximum. Consequently
n
X n
X
Sb (ρ||σ) = λj logb (λj ) − λj logb (νj )
j=1 j=1
Xn Xn
≥ λj logb (λj ) − λj logb (αλj )
j=1 j=1
= − logb (α) ≥ 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 218

218 Problems and Solutions

since 0 ≤ α ≤ 1.

Problem 7. Show that for the quantum relative entropy

Sb (ρ||σ) := tr(ρ logb (ρ) − ρ logb (σ)) = −Sb (ρ) − tr(ρ logb (σ))

the equality

Sb (ρ||ρA ⊗ ρB ) = Sb (ρA ) + Sb (ρB ) − Sb (ρ)

holds.

Solution 7. From σ = ρA ⊗ ρB we find that the eigenvalues of σ are

the products of eigenvalues λA,j of ρA and λB,k of ρB . The corresponding
(orthonormal) eigenstates are |φA,j i ⊗ |φB,k i composed of the orthonormal
eigenstates of ρA and ρB corresponding to λA,j and λB,k . Consequently

Sb (ρ||ρA ⊗ ρB ) = −Sb (ρ) − tr(ρ logb (ρA ⊗ ρB ))

= −Sb (ρ) − trA 
(trB (ρ logb (ρA ⊗ ρB ))) 
m
X
= −Sb (ρ) − trA ρA logb (λA,j )|φA,j ihφA,j |
j=1
n
!
X
−trB ρB logb (λB,k )|φB,k ihφB,k |
k=1
= Sb (ρA ) + Sb (ρB ) − Sb (ρ).

From Klein’s inequality, S(ρ||ρA ⊗ ρB ) ≥ 0, we obtain the property of

subadditivity for the von Neumann entropy

Sb (ρ) ≤ Sb (ρA ) + Sb (ρB ).

Problem 8. An n × n density matrix ρ is a positive semidefinite matrix

such that tr(ρ) = 1. The nonnegative eigenvalues of ρ are the probabilities
of the physical states described by the corresponding eigenvectors. The
entropy of the statistical state described by the density matrix ρ is defined
by
S(ρ) := −tr(ρ ln(ρ)).
For the n × n hermitian matrix H (Hamilton operator) the statistical av-
erage of the energy E is defined by

E := tr(Hρ).

Let
ψ(ρ) := tr(Hρ) − tr(ρ ln(ρ)).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 219

Entropy 219

(i) Show that ln(tr(eH )) = max { tr(Hρ) + S(ρ) }.

(ii) Show that −S(ρ) = max { tr(Hρ) − ln(tr(eH )) }.

Solution 8. (i) For every n × n hermitian matrix Q, and for each  ∈ R

in a neighbourhood of 0, consider the differentiable function

f () = ψ(e−iQ ρeiQ ).

Since eiQ is a unitary matrix and the trace is invariant under unitary
transformation we have

f () = tr(He−iQ ρeiQ ) − tr(ρ ln(ρ)).

By the extremum condition it follows that

df (0)
= itr(Q[H, ρ]) = 0
d
where [ , ] denotes the commutator. Since Q is arbitrary, we conclude that
[H, ρ] = 0n and therefore eH and ρ also commute. Using tr(ρ) = 1, we
obtain

tr(Hρ) + S(ρ) = tr(ρ(ln(eH ) − ln(ρ)))

= ln(tr(eH )) − tr(ρ(ln(tr(eH )) + ln(ρ) − ln(eH ))).

Since eH and ρ commute, it follows that

tr(Hρ) + S(ρ) = ln(tr(eH )) − tr(eH ρe−H ln(tr(eH )ρe−H )).

Setting C := tr(eH )ρe−H we obtain

ln(tr(eH )) − tr(eH ρe−H ln(tr(eH )ρe−H )) =

ln(tr(eH )) − (treH )−1 tr(eH (C ln(C) − C + In )).

Since x ln(x) − x + 1 ≥ 0 for x ≥ 0, we conclude that the last expression is
less than or equal to ln(tr(eH )) and equality occurs only if C = In . This
means
eH
ρ= .
tr(eH )
If ρ = eH /tr(eH ) we find that ψ(eH /tr(eH )) = ln(tr(eH )).
(ii) Since

tr((H + kIn )ρ) − ln(tr(eH+kIn )) = tr(Hρ) − ln(tr(eH )), k∈R

we may assume that tr(eH ) = 1. Following an argument similar to the one

in (i) we can show that the maximum of tr(Hρ) − ln(tr(eH )) for hermitian
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 220

220 Problems and Solutions

matrices H occurs when [H, ρ] = 0n . Thus [ρ, eH ] = 0n . Since tr(ρ) −

tr(eH ) = 0n , we have

−tr(Hρ) + tr(ρ ln(ρ)) = tr(eH ρe−H ln(ρe−H ) − tr(eH ρe−H ) + tr(eH )

= tr(eH (Z ln(Z) − Z + In ))
≥0

where Z := ρe−H . Hence the maximum occurs when H = ln(ρ).

Problem 9. Consider the normalized states |ψk i, k = 0, 1, . . . , N − 1

in the Hilbert space CN . A positive operator valued measure is specified
by a decomposition of the identity matrix IN into M positive semidefinite
matrices Pm , i.e.
MX−1
IN = Pm .
m=0

The mutual information is defined by

N −1 M −1  
X X pnm
I= pnm logN
n=0 m=0
pn· p·m

where pnm := hψn |Pm |ψn i are the joint probabilities and
M
X −1 N
X −1
pn· := pnm , p·m := pnm
m=0 n=0

are their marginals. Let M = N = 2 and

   
1 1 1 1 1 −1
P0 = , P1 =
2 1 1 2 −1 1
   
1 1 0
|ψ0 i = √ , |ψ1 i = .
2 −1 1
Find pnm , pn· , p·m and then I.

Solution 9. Straightforward calculation yields

1
p00 = hψ0 |P0 |ψ0 i = 0, p10 = hψ1 |P0 |ψ1 i =
2
3
p01 = hψ0 |P1 |ψ0 i = 1, p11 = hψ1 |P1 |ψ1 i = .
2
Thus
1 3
p0· = 1, p1· = 1, p·0 = , p·1 =
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 221

Entropy 221

and I = 0.

Problem 10. The Kullback-Leibler distance between two probability

mass functions
w = (w1 , . . . , wn ), w
e = (w
e1 , . . . , w
en )
is defined by
n  
X wj
D(w || w)
e := wj log .
j=1
w
ej
Show that the Kullback-Leibler distance between two mixtures densities
n
X n
X
wj fj , w
ej fej
j=1 j=1

has the upper bound

 
Xn n
X n
X
D wj fj || ej fej  ≤ D(w || w)
w e + wj D(fj || fej )
j=1 j=1 j=1

with equality if and only if

w f w
ej fej
Pn j j = Pn
j=1 wj fj j=1 w ej fej

for all j.

Solution 10. Utilizing the log-sum inequality we have

    Pn !
n n n
w j fj
Z X
j=1
X X
D wj fj || w
ej fj  = 
e wj fj  log Pn
j=1 j=1 j=1 j=1 w
ej fej
Z Xn
!
wj fj
≤ wj fj log
j=1 w
ej fej
n   n Z !
X wj X fj
= wj log + wj fj log
j=1
w
ej j=1 fej
n
X
= D(w || w)
e + wj D(fj || fej ).
j=1

Problem 11. Let A, B be n × n hermitian matrices acting in the Hilbert

space Cn . Assume that the eigenvalues of A are pairwise different and
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 222

222 Problems and Solutions

analogously for B. Then the normalized eigenvectors |αj i (j = 1, . . . , n) of

A form an orthonormal basis in Cn and analogously for B the normalized
eigenvectors |βj i (j = 1, . . . , n) form an orthonormal basis in Cn . Let
|ψi be a normalized state in Cn . Then there are n possible outcomes for
measurements of each observable and the probabilities pj (A, |ψi), pj (B, |ψi)
(j = 1, . . . , n) are given by

pj (A, |ψi) := |hψ|αj i|2 , pj (B, |ψi) := |hψ|βj i|2 .

Let H|ψi (X) be the Shannon information entropy

n
X
H|ψi (X) := − pj (X, |ψi) ln(pj (X, |ψi))
j=1

corresponding to the probability distribution {pj (X, |ψi)} (j = 1, . . . , n).

The (Maassen-Uffink) entropic uncertainty relation is given by

H|ψi (A) + H|ψi (B) ≥ −2 ln( max |hαj |βk i|) > 0.
1≤j,k≤n

Note that the right-hand side does not involve the state |ψi.
Let
     
0 1 1 0 cos(θ)
A = σ1 = , B = σ3 = , |ψi = .
1 0 0 −1 sin(θ)

Calculate the left and right-hand side of the entropic uncertainty relation.
Is the entropic uncertainty relation tight for this case?

Solution 11. (i) The eigenvalues and eigenvectors of A = σ1 are given

by    
1 1 1 1
+1 7→ |α1 i = √ , −1 7→ |α2 i = √ .
2 1 2 −1
The eigenvalues and eigenvectors of B = σ3 are given by
   
1 0
+1 7→ |β1 i = , −1 7→ |β2 i = .
0 1

Thus
1 1 1 1
|hα1 |β1 i| = √ , |hα1 |β2 i| = √ , |hα2 |β1 i| = √ , |hα2 |β2 i| = √
2 2 2 2
and the right-hand side of the inequality becomes
 
1
−2 ln( max |hαj |βk i|) = −2 ln √ = ln(2).
1≤j,k≤n 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 223

Entropy 223

Now
1 1
p1 (A, |ψi) = |hα1 |ψi|2 = | cos(θ) + sin(θ)|2 = (1 + sin(2θ))
2 2
2 1 2 1
p2 (A, |ψi) = |hα2 |ψi| = | cos(θ) − sin(θ)| = (1 − sin(2θ))
2 2
and

p1 (B, |ψi) = |hβ1 |ψi|2 = cos2 (θ)

p2 (B, |ψi) = |hβ2 |ψi|2 = sin2 (θ).

Thus
1 1
H|ψi (A) = ln(2) − (1 + sin(2θ)) ln(1 + sin(2θ)) − (1 − sin(2θ)) ln(1 − sin(2θ))
2 2
H|ψi (B) = − cos2 (θ) ln(cos2 (θ)) − sin2 (θ) ln(sin2 (θ)).

Note that with θ = 0 the left-hand side reduces to ln(2), i.e. we have an
equality.

Programming Problems

Problem 1. Consider the density matrix

 
5/12 1/6 1/6
ρ =  1/6 1/6 1/6  .
1/6 1/6 5/12
Show that we have a mixed state. Find the von Neumann entropy.

Solution 1. We evaluate ρ2 and show that ρ2 6= ρ. Then we calculate

the eigenvalues and the von Neumann entropy.
/* mixed.mac */
rho: matrix([5/12,1/6,1/6],[1/6,1/6,1/6],[1/6,1/6,5/12]);
rho2: rho . rho;
if rho=rho2 then print("pure state")
else print("mixed state");
R: eigenvalues(rho);
R: part(R,1);
x1: part(R,1); x2: part(R,2); x3: part(R,3);
if x1>0 then t1: x1;
if x2>0 then t2: x2;
if x3>0 then t3: x3;
t: -t1*log(t1)-t2*log(t2)-t3*log(t3);
t: ratsimp(t);
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 224

224 Problems and Solutions

The eigenvalues are

√ √
1 9 + 57 9 − 57
, , .
4 24 24

Problem 2. Consider the two 3 × 3 hermitian matrices

   
0 0 −i 1 0 0
K1 =  0 0 0  , K2 =  0 0 0 
i 0 0 0 0 −1

and the normalized state in C3

1 T
ψ = √ (1 1 1) .
3
Find the left and right-hand side of the entropic inequality.

Solution 2. The eigenvalues of K1 are −1, 0 +1 we the corresponding

normalized eigenvectors
     
1 0 1
1   1
k11 = √ 0 , k12 =  1  , k13 = √  0  .
2 −i 0 2 i

The eigenvalues of K2 are −1, 0 +1 we the corresponding normalized eigen-

vectors      
1 0 0
k21 =  0  , k22 =  1  , k23 =  0  .
0 0 1
Now we apply the Maxima program
/* entropic.mac */
k11: matrix([1],[0],[-%i])/sqrt(2);
k11T: transpose(k11); k11TC: conjugate(k11T);
k12: matrix([0],[1],[0]);
k12T: transpose(k12); k12TC: conjugate(k12T);
k13: matrix([1],[0],[%i])/sqrt(2);
k13T: transpose(k13); k13TC: conjugate(k13T);
k21: matrix([1],[0],[0]);
k21T: transpose(k21); k21TC: conjugate(k21T);
k22: matrix([0],[1],[0]);
k22T: transpose(k22); k22TC: conjugate(k22T);
k23: matrix([0],[0],[1]);
k23T: transpose(k23); k23TC: conjugate(k23T);
sc1121: abs(k11TC . k21); sc1122: abs(k11TC . k22);
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 225

Entropy 225

sc1123: abs(k11TC . k23); sc1221: abs(k12TC . k21);

sc1222: abs(k12TC . k22); sc1223: abs(k12TC . k23);
sc1321: abs(k13TC . k21); sc1322: abs(k13TC . k22);
sc1323: abs(k13TC . k23);
m: max(sc1121,sc1122,sc1123,sc1221,sc1222,
sc1223,sc1321,sc1322,sc1323);
psi: matrix([1],[1],[1])/sqrt(3);
psiT: transpose(psi); psiTC: conjugate(psiT);
p11: (abs(psiTC . k11))^2;
p12: (abs(psiTC . k12))^2;
p13: (abs(psiTC . k13))^2;
p21: (abs(psiTC . k21))^2;
p22: (abs(psiTC . k22))^2;
p23: (abs(psiTC . k23))^2;
LHS: -p11*log(p11)-p12*log(p12)-p13*log(p13)-p21*log(p21)
-p22*log(p22)-p23*log(p23);
RHS: -2*log(m);

The output 0 for the right-hand side which is obvious since K1 and K2
have a common eigenvector, namely (0 1 0). For the left-hand side we find
2 log(3).

8.3 Supplementary Problems

Problem 1. Consider the Hilbert space Cn . Let A, B be two hermitian

n × n matrices (observable). Assume that A and B have non-degenerate
eigenvalues with the corresponding normalized eigenvectors |a1 i, |a2 i, . . . ,
|an i and |b1 i, |b2 i, . . . , |bn i, respectively. The entropic uncertainty relation
is an inequality given by

S(A) + S(B) ≥ S(AB)

where
n
X n
X
S(A) = − |hψ|aj i|2 ln(|hψ|aj i|2 ), S(B) = − |hψ|bj i|2 ln(|hψ|bj i|2 ),
j=1 j=1

and S(AB) is a positive constant which gives the lower bound of the right-
hand side of the inequality. Consider the Hilbert space C2 . Let
 
cos(θ)
A = σ1 , B = σ2 , |ψi = .
sin(θ)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 226

226 Problems and Solutions

Find S(A) , S(B) and S(A) + S(B) .

Problem 2. Consider the Hilbert space Cn and |ψi ∈ Cn . Let A and

B n × n hermitian matrices (observable) with non-degenerate eigenvalues
and corresponding normalized eigenvectors |uj i, |vj i (j = 1, . . . , n). The
entropic uncertainty relation is an inequality of the form

S(A) + S(B) ≥ SAB

where
n
X n
X
S(A) = − |hψ|uj i|2 ln(|hψ|uj i|2 ), S(A) = − |hψ|vj i|2 ln(|hψ|vj i|2 )
j=1 j=1

and SAB is a positive constant providing the lower bound of the right-hand
side of the inequality. Let
   
0 1 1 0
A = σ1 = , B = σ3 =
1 0 0 −1

and  
cos(θ)
|ψi = .
sin(θ)
Calculate S(A) and S(B) .

Problem 3. Consider the 4 × 4 spin matrix

√
0 −i 3 0 0
 
√
1 i 3 0 −2i 0√ 
S2 = 
2 0 2i 0 −i 3

√
0 0 i 3 0

the normalized entangled vector

1
 
1 1 
|ψi =  .
2 1
−1

Find the left and right-hand side of the entropic inequality.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 227

Chapter 9

Measurement

9.1 Introduction
In quantum measurement models we consider what kind of measurements
can be made on quantum systems as well as how to determine the probabil-
ity that a measurement yields a given result. The effect that measurement
has on the state of a quantum system is also important.

The pure states of a quantum system, S, are described by normalized vec-

tors |ψi which are elements of a Hilbert space H that describes S. The
pure states of a quantum mechanical system are rays in a Hilbert space H
(i.e., unit vectors with an arbitrary phase). The concept of a state as a
ray in a Hilbert space leads to the probability interpretation in quantum
mechanics. Given a physical system in the state ψ, the probability that it
is in the state |χi is
|hψ|χi|2 .

Obviously
0 ≤ |hψ|χi|2 ≤ 1.

While the phase of a vector |ψi has no physical significance, the relative
phase of two vectors does. Consider the Schrödinger equation with time
independent Ĥ
d
i~ |ψ(t)i = Ĥ|ψ(t)i
dt

227
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 228

228 Problems and Solutions

and the initial state |ψ(t = 0)i. Then

|ψ(t)i = exp(−iĤt/~)|ψ(t = 0)i

and
|hψ(0)|e−iĤt/~ |ψ(0)i|2

is the probability to find the state |ψ(t)i in the initial state |ψ(0)i.

A positive operator-valued measure (POVM) is a collection

{ Ej : j = 1, 2, . . . , n }

of nonnegative (positive semi-definite) operators, satisfying

n
X
Ej = I
j=1

where I is the identity operator. In other words a partition of unity (iden-

tity operator) by nonnegative operators is called a positive operator-valued
measure (POVM). When a state |ψi is subjected to such a POVM, outcome
j occurs with probability

p(j) = hψ|Ej |ψi.

For example consider a qubit system

1
E1 = |0ih0|, E2 = |1ih1|, |ψi = √ (|0i + |1i).
2

Since h0|0i = h1|1i = 1 and h0|1i = h1|0i = 0 we find

1 1
p(1) = hψ|E1 |ψi = , p(2) = hψ|E2 |ψi = .
2 2

Measurement can be generalized in the sense that an ancilla system (in a

well defined state), identified by the Hilbert space HA , is introduced and
allowed to interact with the quantum system identified by the Hilbert space
H. The ancilla system is subsequently measured, which may disturb the
original system.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 229

Measurement 229

9.2 Solved Problems

Problem 1. Consider the state
r r
1 2 1 2
|ψi = √ |00i + |11i ≡ √ |0i ⊗ |0i + |1i ⊗ |1i
3 3 3 3
and the product state |φi = |11i ≡ |1i ⊗ |1i. Find p := |hφ|ψi|2 , i.e. the
probability of finding |ψi in the state |φi.

Solution 1. Since h11|00i = 0 and h11|11i = 1 we obtain p = 2/3.

Problem 2. Consider the hermitian matrix (Hamilton operator)

Ĥ = ~ωS1
where S1 is the 3 × 3 spin-1 matrix
 
0 1 0
1 
S1 = √ 1 0 1.
2 0 1 0

(i) Calculate exp(−iĤt/~).

(ii) Consider the normalized vector in C3
 
1
1
|ψ(0)i = √  1  .
3 1

Calculate |ψ(t)i = exp(−iĤt/~)|ψ(0)i.

(iii) Find the probability of finding ψ(t) in the initial state ψ(0), i.e.
|hψ(t)|ψ(0)i|2 .

Solution 2. (i) We have

   
1 0 1 0 1 0
1 1
S12 =  0 2 0, S13 = √  1 0 1  = S1 .
2 2 0
1 0 1 1 0
Using this result we find
i 1
exp(−iĤt/~) = I3 − √ sin(ωt)A + (cos(ωt) − 1)A2
2 2
1 1
− √i2 sin(ωt) 12 cos(ωt) − 21

2 cos(ωt) + 2
=  − √i2 sin(ωt) cos(ωt) − √i2 sin(ωt) .

1 1 √i sin(ωt) 1 1
2 cos(ωt) − 2 − 2 2 cos(ωt) + 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 230

230 Problems and Solutions

(ii) We obtain

cos(ωt) − √i sin(ωt)
 
1 √2
|ψ(t)i = exp(−iĤt/~)|ψ(0)i = √  cos(ωt) − i 2 sin(ωt)  .
3 cos(ωt) − √i sin(ωt) 2

(iii) We find the probability

1 √ 1
|hψ(t)|ψ(0)i|2 = |3 cos(ωt) − i2 2 sin(ωt)|2 = 1 − sin2 (ωt).
9 9

Problem 3. Consider the states (standard basis)

   
1 0
|0i := , |1i :=
0 1

in the Hilbert space C2 and the Bell state

1
|ψi = √ (|0i ⊗ |1i − |1i ⊗ |0i)
2

in the Hilbert space C4 . Let (α, β ∈ R)

|αi := cos(α)|0i + sin(α)|1i, |βi := cos(β)|0i + sin(β)|1i

be states in C2 . Find the probability

p(α, β) := |(hα| ⊗ hβ|)|ψi|2 .

Discuss p as a function of α and β.

Solution 3. Since h0|0i = h1|1i = 1, h0|1i = h1|0i = 0 it follows that

(h0| ⊗ h1|)(|0i ⊗ |1i) = 1, (h1| ⊗ h0|)(|1i ⊗ |0i) = 1 .

We find
1
(cos(α) sin(β) − sin(α) cos(β))2 .
p(α, β) =
2
Using a trigonometric identity we arrive at
1
p(α, β) = sin2 (α − β).
2

Thus p(α, β) ≤ 1/2 for all α, β since sin2 (φ) ≤ 1 for all φ ∈ R. For example,
if α = β we have p = 0. If α − β = π/2 we have p = 1/2.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 231

Measurement 231

Problem 4. Consider the normalized entangled state

1 1
|ψi = √ (|01i − |10i) ≡ √ (|0i ⊗ |1i − |1i ⊗ |0i)
2 2
and h0| ⊗ I2 , where I2 is the 2 × 2 unit matrix. Find (h0| ⊗ I2 )|ψi. Discuss.

Solution 4. Since h0|0i = 1, h0|1i = 0 and I2 |1i = |1i, we obtain

1
(h0| ⊗ I2 )|ψi = √ |1i.
2
The first system is measured with probability 1/2 and the system collapses
to the state |1i (partial measurement).

Problem 5. Let { |0i, |1i } denote an orthonormal basis in C2 . In other

words
h0|1i = h1|0i = 0, h0|0i = h1|1i = 1.
(i) Show that for α, β ∈ R,

A := α|0ih0| + β|1ih1|

is an observable. Describe the measurement outcomes and associated prob-

abilities when measuring the first qubit of the two qubit system described
by
1
|ψi = √ (|0i ⊗ |0i + |1i ⊗ |1i)
2
where the measurement is described by A.
(ii) Let |φi be the state of the system after the measurement in (i). Describe
the measurement outcomes and associated probabilities when measuring
the second qubit of A.

Solution 5. (i) To show that A is an observable it suffices to show that

A∗ = A. Using

(|0ih0|)∗ = h0|∗ |0i∗ = |0ih0|, (|0ih1|)∗ = h1|∗ |0i∗ = |1ih0|,

(|1ih0|)∗ = h0|∗ |1i∗ = |0ih1|, (|1ih1|)∗ = h1|∗ |1i∗ = |1ih1|

we find that

A∗ = (α|0ih0| + β|1ih1)∗ = (α|0ih0|)∗ + (β|1ih1|)∗ = α|0ih0| + β|1ih1|

= α|0ih0| + β|1ih1| = A

since α = α for α ∈ R. Since we work in C2 , A has two eigenvalues. From

A|0i = (α|0ih0| + β|1ih1)|0i = α|0ih0|0i + β|1ih1|0i = α|0i

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 232

232 Problems and Solutions

and

A|1i = (α|0ih0| + β|1ih1)|1i = α|0ih0|1i + β|1ih1|1i = β|1i

the two eigenvalues (i.e. measurement outcomes) are α and β with corre-
sponding orthonormal eigenstates |0i and |1i. Thus for measuring the first
qubit we consider the observable A ⊗ I2 with eigenvalues α (eigenstates
|0i ⊗ |0i and |0i ⊗ |1i) and β (eigenstates |1i ⊗ |0i and |1i ⊗ |1i). This is
not the only choice for the corresponding eigenstates, but is a convenient
one. It will be useful to calculate some scalar products in advance

1
(|0i ⊗ |0i)∗ |ψi = (h0| ⊗ h0|) √ (|0i ⊗ |1i − |1i ⊗ |0i)
2
1
= √ (h0|0i ⊗ h0|1i − h0|1i ⊗ h0|0i) = 0,
2

1 1
(|0i ⊗ |1i)∗ |ψi = √ , (|1i ⊗ |0i)∗ |ψi = − √ , (|1i ⊗ |1i)∗ |ψi = 0.
2 2
We need to consider two possibilities, namely α = β and α 6= β.
For α = β there is only one measurement outcome: α. The corresponding
projection operator onto the eigenspace is determined from the eigenstates

Πα := (|0i ⊗ |0i)(|0i ⊗ |0i)∗ + (|0i ⊗ |1i)(|0i ⊗ |1i)∗

+(|1i ⊗ |0i)(|1i ⊗ |0i)∗ + (|1i ⊗ |1i)(|1i ⊗ |1i)∗
= |0ih0| ⊗ |0ih0| + |0ih0| ⊗ |1ih1| + |1ih1| ⊗ |0ih0| + |1ih1| ⊗ |1ih1|
= I2 ⊗ I2 = I4

i.e. the identity operator. The probability of obtaining the measurement

outcome α is
1
pα = kΠα |ψik2 = k|ψik2 = hψ|ψi = √ (h0| ⊗ h0| − h1| ⊗ h1|)|ψi
2
 
1 1
= − − = 1.
2 2

The state after measurement is the projected state Πα |ψi = I4 |ψi = |ψi
which when normalized yields |φi = |ψi since hψ|ψi = 1.
α 6= β: Measurement outcome α: The corresponding projection operator
onto the eigenspace is determined from the eigenstates

Πα := (|0i ⊗ |0i)(|0i ⊗ |0i)∗ + (|0i ⊗ |1i)(|0i ⊗ |1i)∗

= |0ih0| ⊗ |0ih0| + |0ih0| ⊗ |1ih1|
= |0ih0| ⊗ I2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 233

Measurement 233

Note that
1
Πα |ψi = √ (|0ih0| ⊗ I2 )(|0i ⊗ |1i − |1i ⊗ |0i)
2
1 1
= √ (|0ih0|0i) ⊗ |1i − (|0ih0|1i) ⊗ |0i = √ |0i ⊗ |1i.
2 2
The probability of obtaining the measurement outcome α is
1 1 1
pα = kΠα |ψik2 = (Πα |ψi)∗ Πα |ψi = √ (h0| ⊗ h1|) √ (|0i ⊗ |1i) = .
2 2 2

The state after measurement is the projected state Πα |ψi = √1 |0i ⊗ |1i
2
which when normalized yields
Πα |ψi Πα |ψi
|φi = = √ = |0i ⊗ |1i.
kΠα |ψik pα

Measurement outcome β: The corresponding projection onto the eigenspace

is determined from the eigenstates

Πβ := (|1i ⊗ |0i)(|1i ⊗ |0i)∗ + (|1i ⊗ |1i)(|1i ⊗ |1i)∗

= |1ih1| ⊗ I2 .

Note that Πβ |ψi = − √12 |1i ⊗ |0i. The probability of obtaining the measure-
ment outcome β is
1
pβ = kΠβ |ψik2 = .
2
The state after measurement is the projected state
1
Πβ |ψi = √ |1i ⊗ |0i
2
which when normalized yields
Πβ |ψi Πα |ψi
|φi = = √ = −|1i ⊗ |0i.
kΠβ |ψik pβ

(ii) For measuring the second qubit we consider the observable I2 ⊗ A with
eigenvalues α (eigenstates |0i ⊗ |0i and |1i ⊗ |0i) and β (eigenstates |0i ⊗ |1i
and |1i ⊗ |1i). This is not the only choice for the corresponding eigenstates,
but is a convenient one. The measurement of the second qubit depends on
the results of the first measurement. Thus we need to consider three cases
α = β, and the two outcomes α and β when α 6= β.
α = β: We have
1
|φi = |ψi = √ (|0i ⊗ |1i − |1i ⊗ |0i).
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 234

234 Problems and Solutions

There is only one measurement outcome: α. The corresponding projection

onto the eigenspace is determined from the eigenstates

Πα := (|0i ⊗ |0i)(|0i ⊗ |0i)∗ + (|1i ⊗ |0i)(|1i ⊗ |0i)∗

+(|0i ⊗ |1i)(|0i ⊗ |1i)∗ + (|1i ⊗ |1i)(|1i ⊗ |1i)∗
= |0ih0| ⊗ |0ih0| + |1ih1| ⊗ |0ih0| + |0ih0| ⊗ |1ih1| + |1ih1| ⊗ |1ih1|
= I2 ⊗ I2 = I4

i.e. the identity operator. As above, the probability of the measurement

outcome α is pα,α = 1.
α 6= β: First measurement outcome was α: We have |φi = |0i ⊗ |1i.
Measurement outcome α: The corresponding projection onto the eigenspace
is determined from the eigenstates

Πα,α := (|0i ⊗ |0i)(|0i ⊗ |0i)∗ + (|1i ⊗ |0i)(|1i ⊗ |0i)∗ = I2 ⊗ |0ih0|.

Note that Πα |φi = 0. The probability of obtaining the measurement out-

come α is pα,α = 0.
Measurement outcome β: The corresponding projection onto the eigenspace
is determined from the eigenstates

Πα,β = I2 ⊗ |1ih1|.

Note that Πα,β |φi = |0i ⊗ |1i = |φi. The probability of obtaining the
measurement outcome β is pα,β = 1.
First measurement outcome was β: We have

|φi = −|1i ⊗ |0i.

Measurement outcome α: The corresponding projection onto the eigenspace

is determined from the eigenstates

Πβ,α = I2 ⊗ |0ih0|.

Note that Πα |φi = −|1i ⊗ |0i = |φi. The probability of obtaining the
measurement outcome α is pβ,α = 1.
Measurement outcome β: The corresponding projection operator onto the
eigenspace is determined from the eigenstates

Πβ,β = I2 ⊗ |1ih1|.

Note that Πβ,β |φi = 0. The probability of obtaining the measurement out-
come β is pβ,β = 0.

Tabulating the probabilities for α 6= β we find

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 235

Measurement 235

Outcomes α, α α, β β, α β, β
1 1
Probability pα pα,α = 0 pα pα,β = 2 pβ pβ,α = 2 pβ pβ,β = 0
Consequently, for α 6= β, the probability that the two measurement out-
comes are the same is 0 (impossible) and the probability that the two
measurement outcomes are different is 1 (certain).

Problem 6. Assume that Alice operates a device that prepares a quan-

tum system and Bob does subsequent measurement on the system and
records the results. The preparation device indicates the state the system
is prepared in. A preparation readout event j, where j = 1, 2, . . . , m of the
preparation device is associated with a linear non-negative definite operator
Λj acting on the state space of the system. The operators Λj need not be
orthogonal to each other. The measurement device has a readout event k,
where k = 1, 2, . . . , n that shows the result of the measurement. A mea-
surement device is associated with a measurement device operator Γk which
is also linear and non-negative definite. For a von Neumann measurement
this operator would be a pure state projector. Let
m
X n
X
Λ := Λj , Γ := Γk .
j=1 k=1

Give an interpretation of the following probabilities

tr(Λj Γk )
p(j, k) = (1)
tr(ΛΓ)

tr(Λj Γ)
p(j) = (2)
tr(ΛΓ)
tr(ΛΓk )
p(k) = (3)
tr(ΛΓ)
tr(Λj Γk )
p(k|j) = (4)
tr(Λj Γ)
tr(Λj Γk )
p(j|k) = . (5)
tr(ΛΓk )

Solution 6. Expression (1) is the probability associated with a particular

point (j, k) in the sample space. Expression (2) is the probability that, if
an experiment chosen at random has a recorded combined event, this event
includes preparation event j. Expression (3) is the probability that the
recorded combined event includes the measurement event k. Expression
(4) is the probability that, if the recorded combined event includes event j,
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 236

236 Problems and Solutions

it also includes event k. Thus it is the probability that the event recorded by
Bob is the detection of the state corresponding to Γk if the state prepared
by Alice in the experiment corresponds to Λj . This expression can be used
for prediction. To calculate the required probability from the operator Λj
associated with the preparation event j, every possible operator Γk must
be known, that is, the mathematical description of the operation of the
measuring device must be known. Analogously, (5) is the probability that
the state prepared by Alice corresponds to Λj if the event recorded by Bob
is the detection of the state corresponding to Γk . This expression can be
used for retrodiction if Γk and all the Λj operators of the preparation device
are known.

Problem 7. Let A be an n × n hermitian matrix. Then the eigenvalues

λj , j = 1, 2, . . . , n are real. Assume that all eigenvalues are distinct. The
matrix A can be written as (spectral representation)
n
X
A := λ j Pj , Pj := |uj ihuj | (1)
j=1

where |uj i are the normalized eigenvectors of A with eigenvalue λj . For

the projectors Pj we have Pj Pk = δjk Pj . Every observable A defines a pro-
jective measurement. A state |ψi in Cn subject to projective measurement
by observable (1) goes into the state
P |ψi
p j
hψ|Pj |ψi

with probability

p(j) = hψ|Pj |ψi ≡ hψ|uj ihuj |ψi = |hψ|uj i|2 .

The eigenvalues λj are registered as the measured value. If the system is

subjected to the same measurement immediately after a projective mea-
surement, the same outcome occurs with certainty. The expectation of the
measured value is
Xn
hAi = λj p(j) = hψ|A|ψi.
j=1

(i) Let  
0 −i
A = σ2 = .
i 0
Find the spectral representation of A.
(ii) Let  
1 1
|ψi = √ .
2 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 237

Measurement 237

Calculate the probabilities corresponding to the eigenvalues λ1 and λ2 of A

p(λ1 ) = hψ|Pλ1 |ψi, p(λ2 ) = hψ|Pλ2 |ψi.

Solution 7. (i) The eigenvalues of A are λ1 = 1 and λ2 = −1. The

corresponding eigenvectors are
   
1 1 1 1
|u1 i = √ , |u2 i = √ .
2 i 2 −i

Thus    
1 1 1 1 −i
Pλ1 = |u1 ihu1 | = (1 − i) =
2 i 2 i 1
   
1 1 1 1 i
Pλ2 = |u2 ihu2 | = (1 i) =
2 −i 2 −i 1
with I2 = Pλ1 + Pλ2 and A = Pλ1 − Pλ2 .
(ii) We have
     
1 1 −i 1 1 1 1−i
Pλ1 |ψi = √ = √
2 i 1 2 1 2 2 1+i
     
1 1 i 1 1 1 1+i
Pλ2 |ψi = √ = √ .
2 −i 1 2 1 2 2 1−i
Thus
1 1
p(λ1 ) = hψ|Pλ1 |ψi = , p(λ2 ) = hψ|Pλ2 |ψi = .
2 2

Problem 8. Let (θ ∈ R)

P (θ) := eiθ |0ih0| + e−iθ |1ih1| ≡ eiθ (|0ih0| + e−i2θ |1ih1|)

denote the phase change transform on a single qubit.

(i) Calculate (φ ∈ R)
   
π φ θ
|s(θ, φ)i := P − UH P UH |0i.
4 2 2

(ii) Determine the probability that the state |s(θ, φ)i is in the state

(a) |0i, (b) |1i, (c) |s(θ0 , φ0 )i.

The real parameters θ and φ can be interpreted as spherical co-ordinates

which define any qubit on the unit sphere called the Bloch sphere.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 238

238 Problems and Solutions

Solution 8. (i) We have

   
π φ θ
|s(θ, φ)i = P − UH P UH |0i
4 2 2
   
π φ θ 1
=P − UH P √ (|0i + |1i)
4 2 2 2
 
π φ 1
=P − UH √ (eiθ/2 |0i + e−iθ/2 |1i)
4 2 2
  
π φ 1 iθ/2 
=P − e (|0i + |1i) + e−iθ/2 (|0i − |1i)
4 2 2
 
π φ
=P − (cos(θ/2)|0i + i sin(θ/2)|1i)
4 2
π
 φ φ

= ei 4 e−i 2 cos(θ/2)|0i + ei 2 sin(θ/2)|1i
π φ
= ei( 4 − 2 ) cos(θ/2)|0i + eiφ sin(θ/2)|1i .

The most general state of a single qubit is described by three real parame-
ters θ, φ, σ ∈ R
eiσ cos(θ/2)|0i + eiφ sin(θ/2)|1i .

The parameter σ represents the global phase, and can be ignored since it
cannot be detected in the measurement model. The same applies to the
global phase exp(i(π/4 − φ/2)) in the derivation. Thus θ and φ can be used
to define any single qubit |s(θ, φ)i.
(ii) For the probabilities (a) we have

|h0|s(θ, φ)i|2 = cos2 (θ/2).

For the probability (b) we have

|h1|s(θ, φ)i|2 = sin2 (θ/2).

For the probability |hs(θ0 , φ0 )|s(θ, φ)i|2 we find

0 2
 i 0 i

|hs(θ0 , φ0 )|s(θ, φ)i|2 = cos(θ/2) cos(θ0 /2)e 2 (φ −φ) + sin(θ/2) sin(θ0 /2)e 2 (φ−φ ) 


where we used h0|0i = h1|1i = 1 and h0|1i = h1|0i = 0. It follows that

|hs(θ0 , φ0 )|s(θ, φ)i|2 =

cos2 ((φ0 − φ)/2) cos2 ((θ0 − θ)/2) + sin2 ((φ0 − φ)/2) cos2 ((θ0 + θ)/2).
If θ0 = θ and φ0 = φ we find 1 for the probability.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 239

Measurement 239

Problem 9. Consider the finite-dimensional Hilbert space Cn with n > 2.

Consider an orthonormal basis

{ |0i, |1i, . . . , |n − 1i } .

Let E be any projector in it, and Ej := |jihj|, where j = 0, 1, . . . , n − 1.

Let the probability of obtaining 1 when measuring E be P (E). Then

P (I) = 1, 0 ≤ P (E) ≤ 1, P (0) = 0, Ej Ek = δjk Ej .

P (E0 + E1 + · · · + En−1 ) = P (E0 ) + P (E1 ) + · · · + P (En−1 ). (1)

A state s is determined by the function P (E) which satisfies (1). Gleason’s
theorem states that for any P (E) which satisfies (1) there exists a density
matrix ρ such that
P (E) = tr(ρE).
In other words, s is described by the density matrix ρ. Show that Gleason’s
theorem does not hold in two-dimensional Hilbert spaces.

Solution 9. In the two-dimensional Hilbert space consider the eigenvalue

equation
(σ · n)|mi = |mi
where σ · n := σ1 n1 + σ2 n2 + σ3 n3 , n is the unit vector in R3 (knk = 1)
with the parameter representation

n := (sin(θ) cos(φ), sin(θ) sin(φ), cos(θ))

where 0 ≤ θ ≤ π, 0 ≤ φ < 2π and

 
cos(θ/2)
|mi := .
eiφ sin(θ/2)

The projector onto |mi is given by

1 −iφ
cos2 (θ/2)
 
Em ≡ |mihm| = 1 iφ 2e sin(θ)
= E(θ, φ)
2
2 e sin(θ) sin (θ/2)
1
since cos(θ/2) sin(θ/2) ≡ 2 sin(θ). Equation (1) holds with

P (Em + E−m ) = P (Em ) + P (E−m ) = P (I) = 1, Em E−m = 0.

It is not difficult to find probability distribution functions Pm = P (θ, φ)

such that no density matrix ρ exists. An example is

1 cos3 (θ)
P (θ, φ) = + .
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 240

240 Problems and Solutions

Problem 10. Consider the two qubits in the Hilbert space C2

|ψ1 i := cos(θ1 /2)|0i + sin(θ1 /2)eiφ1 |1i

|ψ2 i := cos(θ2 /2)|0i + sin(θ2 /2)eiφ2 |1i.

(i) Find the product state |ψi = |ψ1 i ⊗ |ψ2 i in C4 .
(ii) Consider the qutrit state in the Hilbert space C3
1
|φi = √ (|0i + |1i + |2i).
3
To encode the state |ψ1 i ⊗ |ψ2 i we use the state |φi and perform projec-
tive measurements on the state |φi ⊗ (|ψ1 i ⊗ |ψ2 i) given by the projection
operators P0 , P1 , P2 , P3 acting in the Hilbert space C3 ⊗ C4

P0 := |0ih0| ⊗ (|1i ⊗ |0ih1| ⊗ h0|)

+|1ih1| ⊗ (|0i ⊗ |1ih0| ⊗ h1|) + |2ih2| ⊗ (|1i ⊗ |1ih1| ⊗ h1|)
P1 := |0ih0| ⊗ (|0i ⊗ |1ih0| ⊗ h1|)
+|1ih1| ⊗ (|1i ⊗ |1ih1| ⊗ h1|) + |2ih2| ⊗ (|0i ⊗ |0ih0| ⊗ h0|)
P2 := |0ih0| ⊗ (|1i ⊗ |1ih1| ⊗ h1|)
+|1ih1| ⊗ (|0i ⊗ |0ih0| ⊗ h0|) + |2ih2| ⊗ (|1i ⊗ |0ih1| ⊗ h0|)
P3 := |0ih0| ⊗ (|0i ⊗ |0ih0| ⊗ h0|)
+|1ih1| ⊗ (|1i ⊗ |0ih1| ⊗ h0|) + |2ih2| ⊗ (|0i ⊗ |1ih0| ⊗ h1|).

Find the probability p0 := (hφ| ⊗ hψ|)P0 (|φi ⊗ |ψi).

Solution 10. (i) We have

|ψi = |ψ1 i ⊗ |ψ2 i

= cos(θ1 /2) cos(θ2 /2)|0i ⊗ |0i + sin(θ1 /2)eiφ1 cos(θ2 /2)|1i ⊗ |0i
+ cos(θ1 /2) sin(θ2 /2)eiφ2 |0i ⊗ |1i + sin(θ1 /2)eiφ1 sin(θ2 /2)eiφ2 |1i ⊗ |1i.

(ii) Using (i) we have

1
P0 (|φi ⊗ |ψi) = √ |0i ⊗ sin(θ1 /2)eiφ1 cos(θ2 /2)|1i ⊗ |0i
3
1
+ √ |1i ⊗ cos(θ1 /2) sin(θ2 /2)eiφ2 |0i ⊗ |1i
3
1
+ √ |2i ⊗ sin(θ1 /2)eiφ1 sin(θ2 /2)eiφ2 |1i ⊗ |1i.
3
Then
1
(hφ| ⊗ hψ|)P0 (|φi ⊗ |ψi) = sin(θ1 /2)eiφ1 cos(θ2 /2)hψ|(|1i ⊗ |0i)
3
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 241

Measurement 241

1
+ cos(θ1 /2) sin(θ2 /2)eiφ2 hψ|(|0i ⊗ |1i)
3
1
+ sin(θ1 /2)eiφ1 sin(θ2 /2)eiφ2 hψ|(|1i ⊗ |1i).
3
Since
hψ|(|1i ⊗ |0i) = e−iφ1 sin(θ1 /2) cos(θ2 /2)
hψ|(|0i ⊗ |1i) = e−iφ2 cos(θ1 /2) sin(θ2 /2)
hψ|(|1i ⊗ |1i) = e−iφ1 e−iφ2 sin(θ1 /2) sin(θ2 /2)
we obtain the probability
1
1 − cos2 (θ1 /2) cos2 (θ2 /2)


(hφ| ⊗ hψ|)P0 (|φi ⊗ |ψi) =
3
where we used sin2 (α) + cos2 (α) ≡ 1.

Problem 11. Let B be an observable with k possible measurement out-

comes (bj )
k
X k
X
B= bj Πj , Πj = I
j=1 j=1

where Πj denotes mutually orthogonal projection operators. The measure-

ment of a system described by the density operator ρ yields the following
orthogonal measurement:
1. The outcome (eigenvalue) bj is obtained with probability
p(bj ) = tr(Πj ρ) = tr(Πj ρΠj )
where we used the cyclic invariance of the trace and Π2j = Πj .
2. The expectation value (average measurement value) is given by tr(ρB).
3. The state of the system after measurement is in the measured state of
the system (i.e. the system is projected onto the state corresponding to the
measurement outcome)
Πj ρΠj
ρbj = .
p(bj )
Discuss orthogonal measurement of the W state
0 0 0 0 0 0 0 0
 
0 1 1 0 1 0 0 0
0 1 1 0 1 0 0 0
 
10 0 0 0 0 0 0 0

|W ihW | = 
3 0 1 1 0 1 0 0 0

0 0 0 0 0 0 0 0
 
0 0 0 0 0 0 0 0
 
0 0 0 0 0 0 0 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 242

242 Problems and Solutions

with respect to the observable B given by the diagonal matrix

diag(0, 1, 0, 0, 0, 0, 0, 0) + 2diag(0, 0, 1, 0, 0, 0, 0, 0) + 3diag(0, 0, 0, 0, 1, 0, 0, 0).

Solution 11. The measurement outcomes are b1 = 0, b2 = 1, b3 = 2,

b4 = 3 (eigenvalues) with corresponding projection operators

Π1 = diag(1, 0, 0, 1, 0, 1, 1, 1), Π2 = diag(0, 1, 0, 0, 0, 0, 0, 0),

Π3 = diag(0, 0, 1, 0, 0, 0, 0, 0), Π4 = diag(0, 0, 0, 0, 1, 0, 0, 0).

The probability of the outcome b2 = 1 is given by

1
p2 := tr(Π2 W Π2 ) = .
3

Similarly p1 = 0 and p3 = p4 = 13 . However, the density operators

1
ρ1 := diag(0, 1, 1, 0, 1, 0, 0, 0)
3

0 0 0 0 0 0 0 0
 
0 2 1 0 1 0 0 0
0 1 2 0 1 0 0 0
 
10 0 0 0 0 0 0 0

ρ2 := 
6 0 1 1 0 2 0 0 0

0 0 0 0 0 0 0 0
 
0 0 0 0 0 0 0 0
 
0 0 0 0 0 0 0 0
yield the same probabilities for the measurement outcomes with respect
to the measurement B. For the measurement outcome b2 the state of the
system becomes diag(0, 1, 0, 0, 0, 0, 0, 0), for b3 diag(0, 0, 1, 0, 0, 0, 0, 0) and
for b4 diag(0, 0, 0, 0, 1, 0, 0, 0). It is not possible to obtain the measurement
outcome b1 .

Problem 12. Measurement can be generalized in the sense that an ancilla

system (in a well defined state), identified by the Hilbert space HA , is
introduced and allowed to interact with the quantum system identified by
the Hilbert space H. The ancilla system is subsequently measured, which
may disturb the original system. Let ρ be a density operator on H and B
an observable on HA with k possible measurement outcomes
k
X k
X
B= bj Πj , Πj = I
j=1 j=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 243

Measurement 243

where Πj denotes mutually orthogonal projection operators. Thus I ⊗ B

is an observable on the product Hilbert space H ⊗ HA . Let |bi be a nor-
malized eigenvector of B, and consequently an eigenvector of only one of
the Πj . The generalized measurement of a system described by the density
operator ρ yields the following

1. The system ρ is extended with the ancilla in the normalized pure state
|bi which gives the density operator ρ ⊗ |bihb|.
2. The two systems interact via a unitary operator U , i.e. the system is
transformed according to U (ρ ⊗ |bihb|)U ∗ .
3. The outcome bj is obtained with probability
p(bj ) = tr((I ⊗ Πj )U (ρ ⊗ |bihb|)U ∗ (I ⊗ Πj )).
4. The expectation value (average measurement value) is given by
tr(U (ρ ⊗ |bihb|)U ∗ (I ⊗ B)).
5. The state of the system after measurement is in the measured state of
the system (i.e. the system is projected onto the state corresponding to the
measurement outcome)
(I ⊗ Πj )U (ρ ⊗ |bihb|)U ∗ (I ⊗ Πj )
σbj = .
p(bj )
6. The state of the original system after discarding the ancilla system is
given by
ρbj = trHA (σbj ).
Let φB denote an orthonormal basis for HA of normalized eigenvectors of
B, and so |bi ∈ φB . Thus we write the unitary operator U as
X
U= Ujk ⊗ |jihk|.
|ji,|ki∈φB

Describe generalized measurement in terms of the operators Ujk where

|ji, |ki ∈ φB .

Solution 12. We find that the constraint U U ∗ = U ∗ U = I yields

X X
U ∗U = ∗
Ujk Ulm ⊗ |kihj|lihm| = ∗
Ujk Ujm ⊗ |kihm|
|ji,|ki,|li,|mi∈φB |ji,|ki,|mi∈φB
=I
and
X X
UU∗ = ∗
Ujk Ulm ⊗ |jihk|mihl| = ∗
Ujk Ulk ⊗ |jihl|
|ji,|ki,|li,|mi∈φB |ji,|ki,|li∈φB
= I.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 244

244 Problems and Solutions

Equating to I we find
X X
∗ ∗
Ujk Ujm = δkm I, Ujk Ulk = δjl I.
|ji∈φB |ki∈φB

Consequently we find

(I ⊗ Πj )U (ρ ⊗ |bihb|)U ∗ (I ⊗ Πj )
 
X
∗
= (I ⊗ Πj )  Umb ρUnb ⊗ |mihn| (I ⊗ Πj )
|mi,|ni∈φB
X
∗
= (Umb ρUnb ) ⊗ (Πj |mihn|Πj ).
|mi,|ni∈φB

Applying the definition of ρbj yields

X 1 ∗
ρbj = Umb ρUmb
p(bj )
|mi∈φB ,Πj |mi=|mi

where  
X
∗ 
p(bj ) = tr  Umb ρUmb .
|mi∈φB ,Πj |mi=|mi

When the bj are non-degenerate we find

∗
p(bj ) = tr (Umb ρUmb ).

Thus for a given |bi the measurement is described by the operators Umb for
|mi ∈ φB .

Problem 13. Assuming that the bj are non-degenerate (i.e. that mea-
surement yields maximal information) and that |ji is the eigenvector of Πj
(for the eigenvalue 1) we have k operators Um where m = 1, 2, . . . , k and
the generalized measurement
k
X
∗
1. Um Um = I.
m=1
2. The outcome bj is obtained with probability

p(bj ) = tr Uj ρUj∗ .

3. The state of the system ρ after the measurement outcome bj is given by

1
ρbj = Uj ρUj∗ .
p(bj )
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 245

Measurement 245

Consider the Hilbert space H = C2 and the ancilla Hilbert space HA = C2 .

We construct a generalized measurement described by α, β ∈ C
U0 = α|0ih0| + β|1ih1|, U1 = α|1ih1| + β|0ih0|
where {|0i, |1i} forms an orthonormal basis in C2 and |α|2 + |β|2 = 1.
Construct the generalized measurement for U0 and U1 .

Solution 13. We have

U0∗ U0 + U1∗ U1 = (α|0ih0| + β|1ih1|)(α|0ih0| + β|1ih1|)
+(α|1ih1| + β|0ih0|)(α|1ih1| + β|0ih0|)
= |0ih0| + |1ih1| = I2 .
For α = 0 or β = 0, U0 and U1 describe an orthogonal measurement. To
construct the unitary operator
U := U00 ⊗ |0ih0| + U01 ⊗ |0ih1| + U10 ⊗ |1ih0| + U11 ⊗ |1ih1|
we use the commutators given by
[U0 , U0∗ ] = [U1 , U1∗ ] = [U0 , U1 ] = [U0 , U1∗ ] = 02
and set U00 := U0 , U01 := U1∗ , U10 := U1 , U11 := −U0∗ to satisfy that U
is unitary. One orthogonal measurement that implements the generalized
measurement is described by the projection operators
Π1 = |0ih0|, Π2 = |1ih1|
and the corresponding observable
B := −1Π1 + 1Π2 .
Thus the generalized measurement of ρ, a density operator on H, may
proceed as follows:
1. Perform U on the density matrix ρ ⊗ |0ih0|.
2. Measure the ancillary system with respect to the observable B.

Problem 14. (i) Consider the finite-dimensional Hilbert space Cd . A

symmetric informatically complete positive operator valued measure (SIC-
POVM) consists of d2 outcomes that are subnormalized projection matrices
Πj onto pure states
1
Πj = |ψj ihψj |
d
for j, k = 1, . . . , d2 such that
1 + dδjk
|hψk |ψk i|2 = .
d+1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 246

246 Problems and Solutions

Consider the case d = 2. Show that the normalized vectors

 q √ 
(3 + 3)/6
|ψ1 i =  q √ 
eiπ/4 (3 − 3)/6
 q √ 
(3 + 3)/6
|ψ2 i =  q √ 
−eiπ/4 (3 − 3)/6
 q √ 
eiπ/4 (3 − 3)/6
|ψ3 i =  q √ 
(3 + 3)/6
 q √ 
−eiπ/4 (3 − 3)/6
|ψ4 i =  q √ 
(3 + 3)/6

satisfy this condition.

(ii) Consider the matrices σ1 , −iσ2 , σ3 . Find σ1 |ψ1 i, −iσ2 |ψ1 i, σ3 |ψ1 i.
(iii) Let d = 2 and
Sd :=
d
X X 1 1
|ji⊗|ji⊗hj|⊗hj|+ √ (|ji⊗|ki+|ki⊗|ji)⊗ √ (hj|⊗hk|+hk|⊗hj|)
j=1 k>j=1
2 2

where |1i, |2i denotes the standard basis in C2 , i.e.

   
1 0
|1i = , |2i = .
0 1

Show that
2
d
X 2d
(|ψj i ⊗ |ψj i)(hψj | ⊗ hψj |) = .
j=1
d+1

Solution 14. (i) Since

q √ √ q √ √
( (3 + 3)/6)2 = 1/2 + 3/6, ( (3 − 3)/6)2 = 1/2 − 3/6,
q √ q √ √
(3 + 3)/6 (3 − 3)/6 = 1/ 6
we find that the condition is satisfied.
(ii) We obtain σ1 |ψ1 i = |ψ3 i, −iσ2 |ψ1 i = |ψ4 i, σ3 |ψ1 i = |ψ2 i. Thus using
the Pauli spin matrices we can generate |ψ2 i, |ψ3 i, |ψ4 i from |ψ1 i.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 247

Measurement 247

(iii) We obtain
1 0 0 0
 
0 1/2 1/2 0 
S2 =  .
0 1/2 1/2 0
0 0 0 1
Can one find a SIC-POVM in C4 using the states from (i) and the Kronecker
product?

Problem 15. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Consider the

2 × 2 matrix over the complex numbers
 
3
1 X
Π(n) := I2 + nj σj 
2 j=1

where n := (n1 , n2 , n3 ) (nj ∈ R) is a unit vector, i.e. n21 + n22 + n23 = 1.

(i) Describe the property of Π(n), i.e. find Π∗ (n), tr(Π(n)) and Π2 (n),
where tr denotes the trace. The trace is the sum of the diagonal elements
of a square matrix.
(ii) Find the vector  iφ 
e cos(θ)
Π(n) .
sin(θ)
Discuss.

Solution 15. (i) For the Pauli matrices we have σ1∗ = σ1 , σ2∗ = σ2 ,
σ3∗ = σ3 . Thus Π(n) = Π∗ (n). Since tr(σ1 ) = tr(σ2 ) = tr(σ3 ) = 0 and the
trace operation is linear, we obtain tr(Π(n)) = 1. Since σ12 = σ22 = σ32 = I2
and for the anti-commutators
σ1 σ2 + σ2 σ1 = 02 , σ2 σ3 + σ3 σ2 = 02 , σ3 σ1 + σ1 σ3 = 02
the expression
 2
3 3 3 3
1 X 1 1X 1 XX
Π2 (n) = I2 + nj σj  = I2 + nj σj + nj nk σj σk
4 j=1
4 2 j=1 4 j=1
k=1

simplifies to
3 3
1 1X 1X 2
Π2 (n) = I2 + nj σj + n I2 .
4 2 j=1 4 j=1 j

Using n21 + n22 + n23 = 1 we obtain Π2 (n) = Π(n).

(ii) We find
 iφ   
e cos(θ) 1 (1 + n3 )eiφ cos(θ) + (n1 − in2 ) sin(θ)
Π(n) = .
sin(θ) 2 (n1 + in2 )eiφ cos(θ) + (1 − n3 ) sin(θ)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 248

248 Problems and Solutions

Programming Problems

Problem 1. Let |0i, |1i be the standard basis in C2 . Consider the Bell
state
1
|ψi = √ (|0iA ⊗ |0iB + |1iA ⊗ |1iB )
2
where A refers to Alice and B refers to Bob. Let
   
1 0 0 0
Π0 = |0ih0| = , Π1 = |1ih1| =
0 0 0 1

be two projection matrices with Π0 Π1 = 02 . Measurement of the first qubit

(Alice) provides
1 1 1
p1 (0) = hψ|(Π0 ⊗ I2 )∗ (Π0 ⊗ I2 )|ψi = √ (h0| ⊗ h0|) √ (|0i ⊗ |0i) = .
2 2 2

Hence the post-measurement state |φi is given by

1
 
1 0
|φi = p (Π0 ⊗ I2 )|ψi = |0i ⊗ |0i =   .
p1 (0) 0
0

This state is not entangled. The measurement of qubit two (Bob) will then
result with certainty in the same result

p2 (0) = hφ|(I2 ⊗ Π0 )∗ (I2 ⊗ Π0 )|φi = 1.

Give a Maxima implementation of this calculation.

Solution 1. Owing to the structure of the vectors and matrices the

conjugate complex operation could be avoided.
/* AliceBob.mac */
e1: matrix([1,0]); e1T: transpose(e1);
e2: matrix([0,1]); e2T: transpose(e2);
psi: (kronecker_product(e1,e1) + kronecker_product(e2,e2))/sqrt(2);
psiT: transpose(psi);
I2: matrix([1,0],[0,1]);
Pi0: matrix([1,0],[0,0]); Pi0T: transpose(Pi0);
Pi1: matrix([0,0],[0,1]); Pi1T: transpose(Pi1);
p10: psiT . transpose(kronecker_product(Pi0,I2))
. kronecker_product(Pi0,I2) . psi;
phi: (kronecker_product(Pi0,I2) . psi)/sqrt(p10);
phiT: transpose(phi);
p20: phiT . transpose(kronecker_product(I2,Pi0))
. kronecker_product(I2,Pi0) . phi;
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 249

Measurement 249

Problem 2. Consider the standard basis |0i, |1i, the Bell state
1
|ψi = √ (|0i ⊗ |0i + |1i ⊗ |1i).
2
and the projection matrices
1 1
Πa = (I2 − σ1 ), Πb = (I2 + σ3 ).
2 2
Find
hψ|(Πa ⊗ Πb )|ψi.

Solution 2. A Maxima implementation is

/* POVM.mac */
e0: matrix([1],[0]);
e1: matrix([0],[1]);
I2: matrix([1,0],[0,1]);
sig1: matrix([0,1],[1,0]);
sig2: matrix([0,-%i],[%i,0]);
sig3: matrix([1,0],[0,-1]);
psi: (kronecker_product(e0,e0)+kronecker_product(e1,e1))/sqrt(2);
psiT: transpose(psi);
Pia: (I2 + sig1)/2;
Pib: (I2 - sig3)/2;
R: psiT . kronecker_product(Pia,Pib) . psi;

The output is 1/4.

9.3 Supplementary Problems

Problem 1. Let Ĥ be a hermitian n × n matrix describing the Hamilton

operator and acting in the Hilbert space Cn . Let A, B be n × n hermitian
matrices and |ψi ∈ Cn . One defines (quantum correlation function)
1
Q(|ψi) := hψ|(A(t)B − AB(t) + BA(t) − B(t)A)|ψi
2
where
A(t) = eiĤt/~ Ae−iĤt/~ , B(t) = eiĤt/~ Be−iĤt/~ .
(i) Let  
cos(θ)
Ĥ = ~ωσ2 , A = σ1 , B = σ3 , |ψi = .
sin(θ)
Find Q(|ψi.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 250

250 Problems and Solutions

(ii) Let
Ĥ = ~ωσ2 ⊗ σ2 , A = σ1 ⊗ σ1 , B = σ3 ⊗ σ3 ,
cos(φ1 )
 
sin(φ1 ) cos(φ2 )
|ψi =  .
 
sin(φ1 ) sin(φ2 ) cos(φ3 )
sin(φ1 ) sin(φ2 ) sin(φ3 )
Find Q(|ψi).

Problem 2. Consider
   
1 1 1 1
|ψ1 i = √ , |ψ2 i = √
2 1 2 −1

and the Bell matrix

1 0 0 1
 
1 0 1 1 0 
U=√  .
2 0 1 −1 0
1 0 0 −1

Find

U (|ψ1 i ⊗ |ψ2 i), (hψ1 | ⊗ hψ2 |)U (|ψ1 i ⊗ |ψ2 i), |(hψ1 | ⊗ hψ2 |)U (|ψ1 i ⊗ |ψ2 i)|2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 251

Chapter 10

Entanglement

10.1 Introduction
Entanglement is the characteristic trait of quantum mechanics which en-
forces its entire departure from classical lines of thought. Let H1 and H2
be two finite-dimensional Hilbert spaces and let |ψi ∈ H1 ⊗ H2 . Then |ψi
is said to be disentangled, separable or a product state if there exist states
|ψ1 i ∈ H1 and |ψ2 i ∈ H2 such that

|ψi = |ψ1 i ⊗ |ψ2 i

otherwise |ψi is said to be entangled . For example the normalized state in

C4
1
 
   
1 1 1 1 1 1
 = √ ⊗√
2 1 2 1 2 1
1
is a product state. For example, a polarization entangled state is
1
√ (|Hi ⊗ |V i + eiφ |V i ⊗ |Hi)
2
where H denotes horizontal polarization and V vertical polarization. This
is one of the Bell states.

251
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 252

252 Problems and Solutions

Let |0i, |1i be an orthonormal basis in C2 . One defines the four Bell states
as
1 1
|Φ+ i = √ (|0i ⊗ |0i + |1i ⊗ |1i), |Φ− i = √ (|0i ⊗ |0i − |1i ⊗ |1i),
2 2

1 1
|Ψ+ i = √ (|0i ⊗ |1i + |1i ⊗ |0i), |Ψ− i = √ (|0i ⊗ |1i − |1i ⊗ |0i).
2 2
If we select the standard basis for |0i, |1i, then we have

1 1 0 0
       
1 0 1  0  1 1 1  1 
√  , √  , √  , √  .
2 0 2 0 2 1 2 −1
1 −1 0 0

The Bell states are fully entangled. They also form an orthonormal basis
in C4 .

A density operator is said to be separable if there exists m ∈ N and density

operators ρ1,j , ρ2,j (j = 1, . . . , m) such that
m
X m
X
ρ= pj ρ1,j ⊗ ρ2,j , pj = 1, pj ∈ [0, 1].
j=1 j=1

There are several measures of entanglement, for example the von Neumann
entropy, the tangle and the Schmidt number. An entanglement measure E
has to satisfy several requirements. For example, if the density matrix ρ is
separable then
E(ρ) = 0.

The entanglement of a maximally entangled state of two n-dimensional

systems should be given by log(n). There should be no increase in entan-
glement under LOCC, i.e. local operations to the density matrix ρ and
classically communicating cannot increase the entanglement of ρ. The en-
tanglement measure should be a convex function.

Consider the Hilbert space H = Cn and the product Hilbert space H ⊗ H.

Let A be an arbitrary n × n matrix over C and In the n × n identity matrix.
Consider the following definition. A normalized vector ψ ∈ H ⊗ H is called
maximally entangled, if its reduced density matrix is maximally mixed, i.e.
a multiple of In
hψ|(A ⊗ In )|ψi = dim(H)−1 tr(A).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 253

Entanglement 253

10.2 Solved Problems

Problem 1. Consider the Hilbert space C2 ⊗ C2 ∼
= C4 and the unitary
2 × 2 matrix
e−iφ sin(θ/2)
 
cos(θ/2)
U (θ, φ) := iφ .
−e sin(θ/2) cos(θ/2)
Show that the state in C4
1
 
0

(U (θ1 , φ1 ) ⊗ U (θ2 , φ2 )  
0
0
is not entangled.

Solution 1. We have
1
 
   
0 1 1
 
U (θ1 , φ1 ) ⊗ U (θ2 , φ2 )   = U (θ1 , φ1 ) ⊗ U (θ2 , φ2 ) .
0 0 0
0

Problem 2. Can the Bell state

1 1
√ (|01i − |10i) ≡ √ (|0i ⊗ |1i − |1i ⊗ |0i)
2 2
in the Hilbert space C4 be written as a product state?

Solution 2. This state cannot be written as product state. Assume that

1
(c0 |0i+c1 |1i)⊗(d0 |0i+d1 |1i) = √ (|0i⊗|1i−|1i⊗|0i) , c0 , c1 , d0 , d1 ∈ C
2
where |c0 |2 + |c1 |2 = 1 and |d0 |2 + |d1 |2 = 1. Then we obtain the system of
four equations
1 1
c0 d0 = 0, c0 d1 = √ , c1 d0 = − √ , c1 d1 = 0.
2 2
This set of equations admits no solution. Thus the Bell state cannot be
written as a product state. The Bell state is entangled.

Problem 3. Let |0i, |1i be an arbitrary orthonormal basis. Can the state
in C4
1 1 1 1
|ψi = √ |0i ⊗ |0i + √ |0i ⊗ |1i + √ |1i ⊗ |0i + √ |1i ⊗ |1i
2 8 8 4
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 254

254 Problems and Solutions

be written as a product state?

Solution 3. From (c1 |0i + c2 |1i) ⊗ (d1 |0i + d2 |1i) = |ψi we obtain the
two conditions
1 1
c1 d1 c2 d2 = √ , c1 d1 c2 d2 = .
8 8
Consequently we have a contradiction and the state |ψi cannot be written
as a product state.

Problem 4. Consider the Hilbert space C2 ⊗ C2 and the unitary 2 × 2

matrix
e−iφ sin(θ/2)
 
cos(θ/2)
U (θ, φ) := .
−eiφ sin(θ/2) cos(θ/2)
Show that the state
1
 
 1
0

U (θ1 , φ1 ) ⊗ U (θ2 , φ2 ) √  
2 0
1
is entangled.

Solution 4. We use the fact that the vector (x1 , x2 , x3 , x4 )T ∈ C4 is

separable if and only if x1 x4 = x2 x3 . We obtain

U (θ1 , φ1 )(1, 0)T ⊗ U (θ2 , φ2 )(1, 0)T + U (θ1 , φ1 )(0, 1)T ⊗ U (θ2 , φ2 )(0, 1)T

cos(θ1 /2) cos(θ2 /2) + e−i(φ1 +φ2 ) sin(θ1 /2) sin(θ2 /2)
 
−iφ1 iφ2
 cos(θ2 /2)e sin(θ1 /2) − cos(θ1 /2)e sin(θ/2/2) 
= .
cos(θ1 /2)e−iφ2 sin(θ2 /2) − cos(θ2 /2)eiφ1 sin(θ1 /2)
i(φ1 +φ2 )
cos(θ1 /2) cos(θ2 /2) + e sin(θ1 /2) sin(θ2 /2)
Hence x1 x4 6= x2 x3 and the state is entangled.

Problem 5. Consider the state

1
|ψi = (|0i ⊗ |0i + eiφ1 |0i ⊗ |1i + eiφ2 |1i ⊗ |0i + eiφ3 |1i ⊗ |1i).
2
(i) Let φ3 = φ1 + φ2 . Is the state |ψi a product state?
(ii) Let φ3 = φ1 + φ2 + π. Is the state |ψi a product state?

Solution 5. (i) We have a product state, i.e.

1 1
√ (|0i + eiφ1 |1i) ⊗ √ (|0i + eiφ2 |1i).
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 255

Entanglement 255

(ii) We do not have a product state we have a maximally entangled state.

Problem 6. Can we find 2 × 2 matrices S1 and S2 such that

1
 
   
1 1 1 0
(S1 ⊗ S2 ) ⊗ = √  . (1)
0 0 2 0
1

Solution 6. From (1) we find

1
 
     
1 1 1 0
S1 ⊗ S2 = √  .
0 0 2 0
1
Thus
1
 
 (1)   (2) 
s11 s11 1 0
(1) ⊗ (2) =√  
s21 s21 2 0
1
or
(1) (2)
 
s11 s11 1
 
 (1) (2)  1 0
 s11 s21 
 (1) (2)  = √   .
 s21 s11  2 0
(1) (2)
s21 s21 1

Thus we have the four conditions

(1) (2) 1 (1) (2) (1) (2) (1) (2) 1
s11 s11 = √ , s11 s21 = 0, s21 s11 = 0, s21 s21 = √
2 2
which are not compatible. Thus no S1 and S2 exist such that (1) is satisfied.

Problem 7. In the Hilbert space C4 we can test whether a state is

entangled or not by calculating the von Neumann entropy.
(i) Consider the Hilbert space HA ⊗ HB , where HA = HB = C2 and the
state
1
 
1  −1 
|ψi :=  .
2 −1
1
Calculate the density matrices using the partial trace

ρA := trHB (|ψihψ|), ρB := trHA (|ψihψ|)

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 256

256 Problems and Solutions

and
−tr(ρA log2 (ρA )), −tr(ρB log2 (ρB ))
where −tr(ρA log2 (ρA )) denotes the von Neumann entropy.
(ii) Consider the Bell state

1

1  0 
|ψi := √  .
2 0
−1

Calculate ρA := trHB (|ψihψ|), −tr(ρA log2 (ρA )).

(iii) Consider the state

1


   
1  −1  1 1 1 1 1
|ψi := (U1 ⊗ U2 )  ≡ (U1 ⊗ U2 ) √ ⊗ √
2 −1 2 2 −1 2 −1

1

where U1 and U2 are unitary matrices acting on C2 . Calculate

ρA := trHB (|ψihψ|), −tr(ρA log2 (ρA )).

(iv) Consider the state

1
 
1  0 
|ψi := √ (U1 ⊗ U2 ) 
0

2
−1

where U1 and U2 are unitary matrices acting on C2 . Calculate

ρA := trHB (|ψihψ|), −tr(ρA log2 (ρA )).

Solution 7. (i) We choose the standard basis in C2 to calculate the trace.

For the density matrix ρ we find

1 −1 −1 1
 
1  −1 1 1 −1 
ρ ≡ |ψihψ| =  .
4 −1 1 1 −1
1 −1 −1 1

Therefore
     
1 0 1 0 1
ρA = ⊗ (1 0 ) |ψihψ| ⊗
0 1 0 1 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 257

Entanglement 257
     
1 0 1 0 0
+ ⊗ (0 1 ) |ψihψ| ⊗
0 1 0 1 1
1 0 0 0
   
   
1 0 0 0 0 0 0 1 0 0 1 0
= |ψihψ|  + |ψihψ| 
0 0 1 0 0 1 0 0 0 1 0 0

0 0 0 1
 
1 1 −1
= .
2 −1 1

Analogously
     
1 1 −1 1 1 −1 1 1 −1
ρB = + = .
4 −1 1 4 −1 1 2 −1 1

In this case, ρA = ρB . We diagonalise ρA . The eigenvalues are 0 and

1 with corresponding orthonormal eigenvectors √12 (1, 1)T and √12 (1, −1)T ,
respectively. Thus
    
1 1 1 0 0 1 1
−tr(ρA log2 (ρA )) = −tr
2 1 −1 0 1 1 −1
    
1 1 1 0 0 1 −1
× log2
2 1 −1 0 1 1 1
     
1 1 1 0 log2 0 0 1 −1
= −tr
2 1 −1 0 1 log2 1 1 1
 
0 0
= −tr =0
0 0

where 0 log2 (0) = 0 and 1 log2 (1) = 0. Hence the state |ψi is not entangled.
(ii) We choose the standard basis in C2 to calculate the trace. We have

1 0 0 −1
 
1 0 0 0 0 
ρ = |ψihψ| =  .
2 0 0 0 0
−1 0 0 1

Thus
     
1 0 1 0 1
ρA = ⊗ ( 1 0 ) |ψihψ| ⊗
0 1 0 1 0
     
1 0 1 0 0
+ ⊗ ( 0 1 ) |ψihψ| ⊗
0 1 0 1 1
     
1 0 0 1 1 0 1 1 0
= + = .
2 0 1 2 0 0 2 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 258

258 Problems and Solutions

Therefore
     
1 1 0 1 1 0
−tr(ρA log2 (ρA )) = −tr log2 =1
2 0 1 2 0 1

where log2 (1/2) = −1. The Bell state |ψi is entangled.

(iii) We choose the basis
    
1 0
U2 , U2
0 1

to calculate the partial trace. We have

1 −1 −1 1
 
1  −1 1 1 −1  ∗ ∗
|ψihψ| = (U1 ⊗ U2 )   (U1 ⊗ U2 )
4 −1 1 1 −1
1 −1 −1 1
   
1 1 −1 1 −1
= (U1 ⊗ U2 ) ⊗ (U1∗ ⊗ U2∗ ).
4 −1 1 −1 1

Therefore
     
1 0 1 0 1
ρA = ⊗ (( 1 0 ) U2∗ ) |ψihψ| ⊗ U2
0 1 0 1 0
     
1 0 1 0 0
+ ⊗ (( 0 1 ) U2∗ ) |ψihψ| ⊗ U2
0 1 0 1 1
      
1 1 −1 1 −1 1
= U1 U1∗ ⊗ ( 1 0 )
4 −1 1 −1 1 0
      
1 1 −1 1 −1 0
+ U1 U1∗ ⊗ ( 0 1 )
4 −1 1 −1 1 1
   
1 1 −1 1 1 −1
= U1 U1∗ + U1 U1∗
4 −1 1 4 −1 1
 
1 1 −1
= U1 U1∗ .
2 −1 1

We diagonalise ρA . The eigenvalues are 0 and 1 with corresponding or-

T T
thonormal eigenvectors √12 U1 ( 1 1 ) and √12 U1 ( 1 −1 ) , respectively.
Thus
      
1 1 1 0 0 1 1
−tr(ρA log2 (ρA )) = −tr U1 U1∗ U1 U1∗
2 1 −1 0 1 1 −1
      
1 1 1 0 0 1 −1
× U1 U1∗ log2 U1∗ U1∗
2 1 −1 0 1 1 1
 
0 0
= −tr = 0.
0 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 259

Entanglement 259

(iv) We choose the basis

    
1 0
U2 , U2
0 1

to calculate the partial trace. We have

1 0 0 −1
 
1  0 0 0 0  ∗ ∗
|ψihψ| = (U1 ⊗ U2 )   (U1 ⊗ U2 )
4 0 0 0 0
−1 0 0 1

and therefore
1 0 0 −1 
 
    
1 0  0 0 0 0  1 0 1
U1∗ ρA U1 = ⊗ (1 0) ⊗
0 1 0 0 0 0 0 1 0

−1 0 0 −1
1 0 0 −1 
 
    
1 0  0 0 0 0  1 0 0
+ ⊗ (0 1) ⊗
0 1 0 0 0 0 0 1 1

−1 0 0 −1
     
1 0 0 1 1 0 1 1 0
= + = .
2 0 1 2 0 0 2 0 1

Thus  
1 1 0
ρA = U1 (U1∗ ρA U1 )U1∗ = U1 U1∗ .
2 0 1
We choose the basis     
1 0
U1 , U1
0 1
to calculate the trace. Thus
     
1 1 0 1 0
−tr(ρA log2 (ρA )) = −tr U1 U1∗ log2 (1/2)U1 U1∗
2 0 1 0 1
   
1 log2 (1/2) 0
= −tr U1 U1∗ = 1
2 0 log2 (1/2)

where we used the cyclic invariance of the trace, log2 (1/2) = −1 and that
U1 is a unitary matrix, i.e. U1 U1∗ = I2 .

Problem 8. Let HA and HB be two finite-dimensional Hilbert spaces

over C. Let |ψi denote a pure state in the Hilbert space HA ⊗ HB . Let
{|0i, |1i} denote an orthonormal basis in C2 . The Schmidt number (also
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 260

260 Problems and Solutions

called the Schmidt rank) of |ψi ∈ HA ⊗ HB over HA ⊗ HB is the smallest

non-negative integer Sch(|ψi, HA , HB ) such that |ψi can be written as
Sch(|ψi,HA ,HB )
X
|ψi = |ψj iA ⊗ |ψj iB
j=1

where |ψj iA ∈ HA and |ψj iB ∈ HB . Let

min(d1 ,d2 )
X
|ψi = λj |jiA ⊗ |jiB
j=1

be the Schmidt decomposition of |ψi over HA ⊗ HB , where d1 and d2 are

the dimensions of the subsystems. Then the Schmidt number is the number
of non-zero λj . The λ2j are the eigenvalues of the matrix trB (|ψihψ|). A
separable state has Schmidt number 1 and an entangled state has Schmidt
number greater than 1.
Let f : {0, 1}2 → {0, 1} be a boolean function. We define the state
1 X
|ψf i := (−1)f (a,b) |ai ⊗ |bi. (1)
2
a,b∈{0,1}

For f we select the AND, OR and XOR operations. The AND, OR and
XOR operations are given by
a b AND(a,b) OR(a,b) XOR(a,b)
0 0 0 0 0
0 1 0 1 1
1 0 0 1 1
1 1 1 1 0

Find the Schmidt numbers of |ψAN D i, |ψOR i and |ψXOR i over C2 ⊗ C2 .

Solution 8. From (1) we obtain

1
|ψAN D i = ((−1)0 · 0 |00i + (−1)0 · 1 |01i + (−1)1 · 0 |10i + (−1)1 · 1 |11i)
2
1
= (|00i + |01i + |10i − |11i)
2
where · denotes the AND operation. Analogously we find for the OR and
XOR operations
1
|ψOR i = (|00i − |01i − |10i − |11i)
2
1
|ψXOR i = (|00i − |01i − |10i + |11i).
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 261

Entanglement 261

Next we take the partial trace of |ψAN D ihψAN D |. We obtain

trB (|ψAN D ihψAN D |) = (I2 ⊗ h0|)|ψAN D ihψAN D |(I2 ⊗ |0i)
+(I2 ⊗ h1|)|ψAN D ihψAN D |(I2 ⊗ |1i)
1
= (2|0ih0| + 2|1ih1|)
4
1
= I2 .
2
In the above calculation we used the fact that
(I2 ⊗ h0|)|abihcd|(I2 ⊗ |0i) + (I2 ⊗ h1|)|abihcd|(I2 ⊗ |1i) = δbd |aihc|
where δbd denotes the Kronecker delta and |abi ≡ |ai ⊗ |bi. Similarly we
find
1
trB (|ψOR ihψOR |) = I2
2
1
trB (|ψXOR ihψXOR |) = (|0ih0| − |0ih1| − |1ih0| + |1ih1|)
2
1
= (|0i − |1i)(h0| − h1|).
2
Clearly the eigenvalues of trB (|ψAN D ihψAN D |) and trB (|ψOR ihψOR |) are
1
2 . Thus

Sch(|ψAN D i, C2 , C2 ) = 2, Sch(|ψOR i, C2 , C2 ) = 2.
The eigenvalues of trB (|ψXOR ihψXOR |) are 0 and 1. Thus
Sch(|ψXOR i, C2 , C2 ) = 1.
We note that |ψXOR i = 21 (|0i − |1i) ⊗ (|0i − |1i).

Problem 9. One particularly interesting state in quantum computing

is the Greenberger-Horne-Zeilinger state (GHZ state). This state of three
qubits acts in the Hilbert space C8 and is given by
           
1 1 1 1 0 0 0
|ψi = √ ⊗ ⊗ + ⊗ ⊗ .
2 0 0 0 1 1 1
(i) Find the density matrix ρ = |ψihψ|.
(ii) Let σ0 ≡ I2 , σ1 , σ2 and σ3 be the Pauli spin matrices, where I2 is the
2 × 2 unit matrix. Show that ρ can be written as a linear combination in
terms of Kronecker products of Pauli matrices (including σ0 ), i.e.
3 3 3
1 X X X
ρ= cj ,j ,j σj ⊗ σj2 ⊗ σj3 .
23 j =0 j =0 j =0 1 2 3 1
1 2 3
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 262

262 Problems and Solutions

Solution 9. (i) We find the dual state

1
hψ| = √ (1 0 0 0 0 0 0 1).
2

Thus
1 0 0 0 0 0 0 1
 
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
 
10 0 0 0 0 0 0 0

ρ=  .
2 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
 
0 0 0 0 0 0 0 0
 
1 0 0 0 0 0 0 1

(ii) We find

1
ρ = (I2 ⊗ I2 ⊗ I2 + I2 ⊗ σ3 ⊗ σ3 + σ3 ⊗ I2 ⊗ σ3 + σ3 ⊗ σ3 ⊗ I2
8
+σ1 ⊗ σ1 ⊗ σ1 − σ1 ⊗ σ2 ⊗ σ2 − σ2 ⊗ σ1 ⊗ σ2 − σ2 ⊗ σ2 ⊗ σ1 )

with I8 = I2 ⊗ I2 ⊗ I2 .

Problem 10. Consider a symmetric matrix A over R

a11 a12 a13 a14

 
a a22 a23 a24 
A =  12
a13 a23 a33 a34

a14 a24 a34 a44

and the Bell basis

1 1
  
1 0 1  0 
|Φ+ i = √   , |Φ− i = √  ,

2 0 2 0
1 −1

0 0
  
1 1 1  1 
+
|Ψ i = √   , |Ψ− i = √  .
2 1 2 −1
0 0

The Bell basis forms an orthonormal basis in R4 . Let Ae denote the matrix
A in the Bell basis. What is the condition on the entries aij such that the
matrix A is diagonal in the Bell basis?
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 263

Entanglement 263

Solution 10. Obviously we have e aij = eaji i.e. the matrix A

e is also
symmetric. Straightforward calculation yields
1
a11 = (Φ+ )T AΦ+ =
e (a11 + 2a14 + a44 )
2
1
a12 = (Φ+ )T AΦ− = (a11 − a44 )
e
2
1
a13 = (Φ+ )T AΨ+ = (a12 + a13 + a24 + a34 )
e
2
+ T − 1
a14 = (Φ ) AΨ = (a12 − a13 + a24 − a34 )
e
2
1
a22 = (Φ− )T AΦ− = (a11 − 2a14 + a44 )
e
2
1
a23 = (Φ− )T AΨ+ = (a12 + a13 − a24 − a23 )
e
2
1
a24 = (Φ− )T AΨ− = (a12 − a13 − a24 + a34 )
e
2
1
a33 = (Ψ+ )T AΨ+ = (a22 + 2a23 + a33 )
e
2
1
a34 = (Ψ+ )T AΨ− = (a22 − a33 )
e
2
1
a44 = (Ψ− )T AΨ− = (a22 − 2a23 + a33 ).
e
2
The condition that the matrix A
e should be diagonal leads to

a11 − a44 = 0, a22 − a33 = 0

and a12 = a13 = a24 = a34 = 0 with the entries a14 and a23 arbitrary. Thus
the matrix A has the form
a11 0 0 a14
 
 0 a22 a23 0 
A= .
0 a23 a22 0
a14 0 0 a11

Problem 11. Consider a bipartite qutrit system HA = HB = C3 with

an arbitrary orthonormal basis { |0i, |1i, |2i } in HA and HB , respectively.
(i) Find the antisymmetric subspace H− on H ⊗ HB .
(ii) Find an arbitrary antisymmetric state on H⊗n .

Solution 11. (i) The antisymmetric subspace H− on HA ⊗ HB is defined

as

H− := spanC { |01i − |10i , |12i − |21i , |20i − |02i } ⊂ HA ⊗ HB .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 264

264 Problems and Solutions

(ii) An antisymmetric state on H⊗n is given by

2
X 2
X
|ψi = aj1 ,j2 ,...,jn ;k1 ,k2 ,...,kn|j1 , . . . , jn ; k1 , . . . , kn i
j1 ,j2 ,...,jn =0 k1 ,k2 ,...,kn =0

where
 n 2 n
1 X Y
aj1 ,j2 ,...,jn ;k1 ,k2 ,...,kn := √ bi1 ,i2 ,...,in im jm km
2 i1 ,i2 ,...,in =0 m=1

and  is the Levi-Civita symbol, i.e. ijk = 1 for (i, j, k) = (1, 2, 3) and its
even permutations, and −1 for odd permutations and 0 otherwise.

Problem 12. Let σ2 be the second Pauli spin matrix. Then

0 0 0 −1
 
   
0 −i 0 −i  0 0 1 0 
σ2 ⊗ σ2 = ⊗ = .
i 0 i 0 0 1 0 0
−1 0 0 0

Find the normalized state (γ ∈ R)

1
 
   
iγσ2 ⊗σ2  0  iγσ ⊗σ 1 1
e  ≡e 2 2 ⊗ .
0 0 0
0

Is the state entangled? Discuss.

Solution 12. Straightforward calculations provide

1 2 1 1
eiγσ2 ⊗σ2 = I4 (1 − γ + γ 4 − · · ·) + iσ2 ⊗ σ2 (γ − γ 3 + · · ·)
2! 4! 3!
= I4 cos(γ) + σ2 ⊗ σ2 (i sin(γ))

Consequently
1 cos(γ)
   
0  0
eiγσ2 ⊗σ2   =  .

0 0
0 −i sin(γ)
If γ = π/4, then the state is entangled and if γ = 0 the state is not entangled
(a product state).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 265

Entanglement 265

Problem 13. Consider the density matrix (Werner state) in C4

1−r
ρw := r|φ+ ihφ+ | + I4
4
where |φ+ i = √12 (1, 0, 0, 1)T is the Bell state, and 0 ≤ r ≤ 1.
(i) Find tr(ρw ) and the eigenvalues of ρw .
(ii) Determine the concurrence

C(ρw ) := max { λ1 − λ2 − λ3 − λ4 , 0 }

where λ1 ≥ λ2 ≥ λ3 ≥ λ4 are the eigenvalues of ρw .

Solution 13. (i) We have

(1 + r)/4 0 0 r/2
 
0 (1 − r)/4 0 0
ρw =  .
 
0 0 (1 − r)/4 0
r/2 0 0 (1 + r)/4
Thus tr(ρw ) = 1. The eigenvalues of ρw are (1 + r)/4 + r/2 = (1 + 3r)/4
and (1 − r)/4 with multiplicity 3.
(ii) From (i) it follows that

λ1 − λ2 − λ3 − λ4 = (1 + 3r)/4 − 3(1 − r)/4 = (3r − 1)/2.

The concurrence is

C(ρw ) = max{(3r − 1)/2, 0}.

1
If r = 0 we have C(ρw ) = 0 and if r = 1 then C(ρw ) = 1. For r = 2 we
find C(ρw ) = 14 .

Problem 14. Let ρ be a density matrix over C2 ⊗ C2 = C4 . We define

the entanglement of formation as
|{pk ,|ψk i}|
X
Ef (ρ) := min pj S(trC2 (|ψj ihψj |))
{pk ,|ψk i}
j=0

where {pk , |ψk i} indicates that the minimum should be taken over all mix-
tures which realize ρ. |{pk , |ψk i}| is the number of pure states comprising
the mixture and
S(σ) := −tr(σ log2 (σ))
is the von Neumann entropy. The minimum is taken over all mixtures

{ (p0 , |ψ0 i), (p1 , |ψ1 i), . . . }

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 266

266 Problems and Solutions

which realize ρ where the cardinality of the set is obviously determined by

the mixture and is finite. We can calculate Ef (ρ) from
p !
1 + 1 − C(ρ)2
Ef (ρ) = h
2

where np p p p o
C(ρ) := max λ1 − λ2 − λ3 − λ4 , 0
is the concurrence, λ1 ≥ λ2 ≥ λ3 ≥ λ4 are the eigenvalues of

ρ(σ2 ⊗ σ2 )ρ∗ (σ2 ⊗ σ2 )

and
h(p) := −p log2 (p) − (1 − p) log2 (1 − p)
is the Shannon entropy. Find Ef (ρ) for the Werner state
5 + + 1 1 1
|φ ihφ |+ |φ− ihφ− | + |ψ + ihψ + | + |ψ − ihψ − | = |φ+ ihφ+ |+ I4


ρw :=
8 8 2 8
where |φ+ i = √1 (1, 0, 0, 1)T is a Bell state.
2

Solution 14. We have

3 0 0 2
 
1 0 1 0 0
ρw =  .
8 0 0 1 0
2 0 0 3
Hence
ρew := (σ2 ⊗ σ2 )ρ∗w (σ2 ⊗ σ2 ) = ρw
where
0 0 0 −1
 
 0 0 1 0 
σ2 ⊗ σ2 =  .
0 1 0 0
−1 0 0 0
Thus
13 0 0 12
 
1 0 1 0 0 
ρw (σ2 ⊗ σ2 )ρ∗w (σ2 ⊗ σ2 ) = ρ2w = .

64 0 0 1 0

12 0 0 13
25 1 1 1
The eigenvalues are 64 , 64 , 64
and 64 . The concurrence is
 
5 1 1 1 1
C(ρ) = max − − − ,0 = .
8 8 8 8 4
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 267

Entanglement 267

This result is consistent with solution 11 when r = 12 . Thus Ef (ρ) = 0.1176.

Problem 15. Let HA and HB denote two finite-dimensional Hilbert

spaces. Consider the Hamilton operator
Ĥ = XA ⊗ XB
−1
where the linear operator XA = XA acts on HA and the linear operator
−1
XB = XB acts on HB . Consequently Ĥ = Ĥ −1 . Let |ψi ∈ HA ⊗ HB .
The von Neumann entropy is given by
E(|ψi) := −trA (ρA log2 (ρA ))
where ρA = trB (|ψihψ|). The entanglement capability of Ĥ is defined as
E(Ĥ) := max Γ(t)|t→0
|ψi∈HA ⊗HB

where
dE(exp(−iĤt)|ψ(0)i)
Γ(t) :=
dt
is the state entanglement rate.
(i) Show that
 
Γ(t) = itrA trB ([Ĥ, |ψihψ|] log2 (ρA ))

where [ , ] denotes the commutator.

(ii) Show that an upper bound on Γ(t)|t→0 is given by Γ(t)|t→0 ≤ 1.9123.

Solution 15. Let ρAB (t) := |ψ(t)ihψ(t)| and ρA (t) := trB (ρAB (t)). We
have
ρAB (t) = exp(−iĤt)ρAB (0) exp(iĤt)
and the time evolution of ρAB (t) (von Neumann equation) is given by
dρAB (t)
i = [Ĥ, ρAB (t)].
dt
Thus
dρA (t)
i = trB [Ĥ, ρAB (t)].
dt
It follows that
d d
Γ(t) = − trA (ρA log2 (ρA )) = −trA ( ρA log2 (ρA ))
dt  dt   
dρA d dρA
= −trA log2 (ρA ) + ρA log2 (ρA ) = −trA log2 (ρA )
dt dt dt
= itrA (trB [Ĥ, ρAB ] log2 (ρA ))
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 268

268 Problems and Solutions

since  
d
trA ρA log2 (ρA ) = 0.
dt
Let
Sch(|ψ(0)i)
X p
|ψ(0)i = λj |φj i ⊗ |ηj i
j=1

be a Schmidt decomposition of |ψ(0)i over HA ⊗ HB , where λj > 0 with

Sch(|ψ(0)i)
X
λj = 1
j=1

and { |φ1 i, . . . , |φSch(|ψ(0)i) i }, { |η1 i, . . . , |ηSch(|ψ(0)i) i } are orthonormal sets

of states. Sch(|ψ(0)i) denotes the Schmidt rank of |ψ(0)i over HA ⊗ HB .
Thus
  Sch(|ψ(0)i)
X h i
trB [Ĥ, ρAB (0)] = I ⊗ hηj | Ĥ, ρAB (0) I ⊗ |ηj i
j=1
Sch(|ψ(0)i)
X
= I ⊗ hηj | [XA ⊗ XB , ρAB (0)] I ⊗ |ηj i
j=1
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X p
= λm λn hηn |XB |ηm iXA |φm ihφn |
m=1 n=1
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X p
− λm λn |φm ihφn |XA hηn |XB |ηm i
m=1 n=1
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X p
= λm λn hηn |XB |ηm i[XA , |φm ihφn |]
m=1 n=1

where we used the result

Sch(|ψ(0)i) Sch(|ψ(0)i)
X X p
ρAB (0) = λm λn (|φm i ⊗ |ηm i)(hφn | ⊗ hηn |).
m=1 n=1

Since
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X
ρA (0) = λj |φj ihφj |, log2 (ρA (0)) = log2 (λj |φj ihφj |)
j=1 j=1

we find

Γ(t)|t→0 = itrA (trB [Ĥ, ρAB (0)] log2 ρA (0))

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 269

Entanglement 269

Sch(|ψ(0)i)
X
=i hφj |trB [Ĥ, ρAB (0)] log2 (ρA (0))|φj i
j=1
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X p
=i λm λn log2 λn hηn |XB |ηm ihφn |XA |φm i
m=1 n=1
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X p
−i λm λn log2 λm hηn |XB |ηm ihφn |XA |φm i
m=1 n=1
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X p λn
=i hηn |XB |ηm ihφn |XA |φm i
λm λn log2
m=1 n=1
λm
Sch(|ψ(0)i) Sch(|ψ(0)i)  
X X p  λn 
≤ λm λn log2 |hηn |XB |ηm i||hφn |XA |φm i|
m=1 n=1
λm 
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X
= (λm + λn )|hηn |XB |ηm i||hφn |XA |φm i|
m=1 n=1
r  
λn λm log2 λn 
 
×
λm + λn λm + λn  λm 
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X
≤ (λm + λn )|hηn |XB |ηm i||hφn |XA |φm i|
m=1 n=1
 
p x
× max x(1 − x) log2
x∈(0,1) 1−x
 
p x
≤ 2 max x(1 − x) log2
x∈(0,1) 1−x
≈ 1.9123
where we used
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X
λm |hηn |XB |ηm i||hφn |XA |φm i|
m=1 n=1
Sch(|ψ(0)i) Sch(|ψ(0)i)
X X
≤ |hηn |XB |ηm i||hφn |XA |φm i| ≤ 1
m=1 n=1

2 2
since XA = I and XB = I.

Problem 16. Consider the Hamilton operator

Ĥ = µ1 σ1 ⊗ σ1 + µ2 σ2 ⊗ σ2 , µ1 , µ2 ∈ R
where σ1 and σ2 are Pauli spin matrices.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 270

270 Problems and Solutions

(i) Calculate the eigenvalues and eigenvectors of Ĥ. Are the eigenvectors
entangled?
(ii) Let |ψi ∈ C4 . The von Neumann entropy is given by

E(|ψi) := −tr(ρA log2 (ρA ))

where ρA := trC2 (|ψihψ|). The entanglement capability of Ĥ is defined as

E(Ĥ) := max4 Γ(t)|t→0

|ψi∈C

where
dE(exp(−iĤt)|ψ(0)i)
Γ(t) :=
dt
is the state entanglement rate. Show that E(Ĥ) = α(µ1 + µ2 ), where
 
p x
α = 2 max x(1 − x) log2 .
x∈(0,1) 1−x

Solution 16. (i) The matrix representation of the Hamilton operator Ĥ

is given by

0 0 0 µ1 − µ2
 
0 0 µ1 + µ2 0
Ĥ =  .
 
0 µ1 + µ2 0 0
µ1 − µ2 0 0 0

The eigenvalues are µ1 − µ2 with corresponding eigenvector

1
|φ+ i = √ (1, 0, 0, 1)T
2
µ2 − µ1 with corresponding eigenvector
1
|φ− i = √ (1, 0, 0, −1)T
2
µ1 + µ2 with corresponding eigenvector
1
|ψ + i = √ (0, 1, 1, 0)T
2
and −µx − µy with corresponding eigenvector

1
|ψ − i = √ (0, 1, −1, 0)T .
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 271

Entanglement 271

Clearly all four eigenvectors are entangled (Bell basis).

(ii) Consider
0
 
√
x
|ψmax i :=  √ 0
 
−i 1 − x0

0
0 0
   
√ √ 1 1 √ √ 1 1 
= ( x0 − i 1 − x0 )   + ( x0 + i 1 − x0 ) 
2 1 2 −1

0 0
where x0 ∈ (0, 1) satisfies
 
p x0
α = 2 x0 (1 − x0 ) log2 .
1 − x0
Now we have
√0
 
cos(t(µ1 + µ2 )) − 1 √ − x0 sin(t(µ1 + µ2 ))
exp(−iĤt)|ψmax i =  √ .


−i x0 sin(t(µ1 + µ2 )) + 1 − x0 cos(t(µ1 + µ2 ))
0
Defining
√ √
a1 := x0 cos(t(µ1 + µ2 )) − 1 − x0 sin(t(µ1 + µ2 ))
√ √
a2 := x0 sin(t(µ1 + µ2 )) + 1 − x0 cos(t(µ1 + µ2 ))
and
ρmax (t) := exp(−iĤt)|ψmax ihψmax | exp(iĤt)
we find
0 0 0 0
 
 
0 a21 ia1 a2 0 a21 0
ρmax (t) =  , trC2 (ρmax (t)) = .
0 −ia2 a1 a22 0 0 a22
0 0 0 0
Thus
d 2
a1 log2 (a21 ) + a22 log2 (a22 )


Γ(t) = −
dt
 
da1 2 da1 da2 2 da2
= − 4a1 log2 (a1 ) + a1 + 4a2 log2 (a2 ) + a2
dt ln(2) dt dt ln(2) dt
= 4µa1 a2 (log2 (a1 /a2 ))
 
p 1
= 4µ x0 (1 − x0 ) cos(2tµ) + (2x0 − 1) sin(2tµ)
2
√ √
x0 cos(2tµ) − 1 − x0 sin(2tµ)

× log2 √ √
x0 sin(2tµ) + 1 − x0 cos(2tµ)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 272

272 Problems and Solutions

where µ ≡ µ1 + µ2 and we used that a1 and a2 satisfy the system of linear

differential equations with constant coefficients
da1 da2
= −a2 , = a1 .
dt dt
Since Ĥ is asymptotically equivalent to (µ1 + µ2 )σ1 ⊗ σ1 and

E((µ1 + µ2 )σ1 ⊗ σ1 ) ≤ (µ1 + µ2 )α

and using
 
p x0
Γ(0) = (µ1 + µ2 )2 x0 (1 − x0 ) log2 = α(µ1 + µ2 )
1 − x0

we find E(Ĥ) = α(µ1 + µ2 ).

Problem 17. Consider the orthonormal basis { |0i, |1i, . . . , |n − 1i } in

the Hilbert space Cn . We assume in the following that this is the standard
basis. Consider the states (coherent states)
n−1
!1/2 n−1
X X √
|βi = 1− xk |0i + xk eiφk |ki
k=1 k=1

where φk ∈ [0, 2π), 0 ≤ xk ≤ 1 and with the constraints

n−1
X
0 ≤ xj ≤ 1 − xk , j = 1, 2, . . . , n − 2.
k=j+1

The Lebesgue measure is given by

n−1
n! Y
dµ(β) = dxj dφj .
(2π)n−1 j=1

(i) Let n = 4. Then the state |βi is given by

(1 − x1 − x2 − x3 )1/2
 
√ iφ1
x e
|βi =  √ 1 iφ2 .
 
x2 e
√ iφ3
x3 e
Show that this state is normalized.
(ii) Calculate the density matrix ρ = |βihβ|.
(iii) Show that the coherent states |βi satisfy
Z
dµ(β)|βihβ| = I4
Ω
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 273

Entanglement 273

where dµ(β) is the uniform measure given above and Ω the domain for φj
(j = 1, 2, 3) and xk (k = 1, 2, 3) described above. I4 is the 4×4 unit matrix.
This equation is called the resolution of identity and a coherent state must
satisfy this condition.
(iv) Find the reduced density matrix from |βi and a condition for entan-
glement.

Solution 17. (i) Taking the scalar product we have

hβ|βi = (1 − x1 − x2 − x3 ) + x1 + x2 + x3 = 1.

Thus the state is normalized.

(ii) We find the 4 × 4 matrix

ρ = |βihβ| =
√ √ √
d2 d x1 e−iφ1 d x2 e−iφ2 d x3 e−iφ3
 
√
 d x1 eiφ1 √ √ i(φ1 −φ2 ) √ √ i(φ1 −φ3 ) 
 √ x1 x1 x2 e x1 x3 e
 d x2 eiφ2 √ √ i(φ2 −φ1 ) √ √ i(φ2 −φ3 ) 
x1 x2 e x2 x2 x3 e 
√ √ √ i(φ3 −φ1 ) √ √ i(φ3 −φ2 )
d x3 eiφ3 x3 x1 e x3 x2 e x3
where d := (1 − x1 − x2 − x3 )1/2 .
(iii) Since
Z 2π Z 2π Z 2π Z 2π
eiφ = 0, dφ3 dφ2 dφ1 = (2π)3
φ=0 φ3 =0 φ2 =0 φ1 =0

and
Z 1 Z 1−x3 Z 1−x2 −x3
1
dx3 dx2 dx1 =
x3 =0 x2 =0 x1 =0 6
Z 1 Z 1−x3 Z 1−x2 −x3
1
dx3 dx2 dx1 x1 =
x3 =0 x2 =0 x1 =0 24
Z 1 Z1−x3 Z 1−x2 −x3
1
dx3 dx2 dx1 x2 =
x3 =0 x2 =0 x1 =0 24
Z 1 Z1−x3 Z 1−x2 −x3
1
dx3 dx2 dx1 x3 =
x3 =0 x2 =0 x1 =0 24

we find (1).
(iv) Let |0i4 , |1i4 , |2i4 , |3i4 be the standard basis in C4 and |0i2 , |1i2 be the
standard basis in C2 . Then we can write |0i4 = |0i2 ⊗|0i2 , |1i4 = |0i2 ⊗|1i2 ,
|2i4 = |1i2 ⊗ |0i2 and |3i4 = |1i2 ⊗ |1i2 with the coefficients
√
c00 = (1 − x1 − x2 − x3 )1/2 , c01 = x1 eiφ1 ,
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 274

274 Problems and Solutions

√ √
c10 = x2 eiφ2 , c11 = x3 eiφ3
which leads to the 2 × 2 matrix
 √ iφ1 
(1 − x1 − x2 − x3 )1/2 x e
C= √ iφ2 √ 1 iφ3 .
x2 e x3 e

The reduced density matrix is

√ √ √
d x2 e−iφ2 + x1 x3 eiφ1 e−iφ3
 
† 1 − x2 − x3
CC = √ √ √
d x2 eiφ2 + x1 x3 e−iφ1 eiφ3 x2 + x3

where d := (1 − x1 − x2 − x3 )1/2 . We obtain

√
det(CC † ) = x3 d2 + x1 x2 − 2 x1 x2 x3 d cos(φ1 + φ2 − φ3 ).

The state |βi is not entangled if det(C † C) = 0.

Problem 18. Consider the pure state

|ψi := α|00i + β|11i

in the Hilbert space C2 ⊗ C2 , where α, β ∈ C and |α|2 + |β 2 | = 1. Let

ρ := |ψihψ| be the corresponding density matrix.
(i) Find −tr(ρ1 log2 (ρ1 )), where ρ1 := trC2 (ρ).
(ii) Let ρe be a density matrix for a disentangled state on C2 ⊗ C2 . Find the
fidelity (also called Uhlmann’s transition probability)
 q 2
√ √
F(ρ, ρe) := tr ρeρ ρ .

(iii) Show that the minimum over ρe of the modified Bures metric

DB (ρ, ρe) := 2 − 2F(ρ, ρe)

is given by 4|α|2 (1 − |α|2 ) at σ := |α|2 |00ih00| + |β|2 |11ih11|. The Bures

metric is defined as
p
DBures (ρ, ρe) := 2 − 2 F(ρ, ρe).

(iv) Compare the result in (iii) with the result from (i).

Solution 18. (i) We find that

ρ = |α|2 |00ih00| + |β|2 |11ih11| + αβ|00ih11| + βα|11ih00|.

Taking the partial trace over the first qubit in C2 yields

ρ1 = trC2 (ρ) = |α|2 |0ih0| + |β|2 |1ih1|.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 275

Entanglement 275

Thus

−tr(ρ1 log2 (ρ1 )) = −|α|2 log2 (|α|2 ) − (1 − |α|2 ) log2 (1 − |α|2 ).

√
(ii) Since ρ is a pure state we have ρ = ρ and
 q 2 h p
√ √ i2
F(ρ, ρe) = tr ρ ρ = tr ρe
ρe ρρ
h p i2 h p i2
= tr |ψihψ|e
ρ|ψihψ| = tr hψ|e
ρ|ψi|ψihψ|
2
= |hψ|e
ρ|ψi| (tr(ρ)) = |hψ|e
ρ|ψi|.

(iii) From (ii) we have

DB (ρ, σ) = 2 − 2F(ρ, σ) = 2 − 2|hψ|σ|ψi|.

For σ = |α|2 |00ih00| + |β|2 |11ih11| we find

DB (ρ, σ) = 2 − 2(|α4 | + |β|4 ) = 2 − 2(|α4 | + (1 − |α|2 )2 ) = 4|α|2 (1 − |α|2 ).

Obviously σ is not entangled. For |α|2 = 0 or |α|2 = 1 it is immediately

clear that we have a minimum. Thus consider 0 < |α|2 < 1. Now let ν be
any fixed density matrix in C4 and λ ∈ [0, 1]. Thus the convex function

σ(λ) := λσ + (1 − λ)ν

is also a density matrix. It follows that

d  d 
DB (ρ, σ(λ)) = −2 |λhψ|σ|ψi + (1 − λ)hψ|ν|ψi| 
 
dλ λ=1 dλ λ=1
d  4 4
 
= −2 λ(|α| + |β| ) + (1 − λ)hψ|ν|ψi 
 dλ 4 λ=1
−2(|α| + |β|4 − hψ|ν|ψi) |α|4 + |β|4 ≥ 0
=
+2(|α|4 + |β|4 − hψ|ν|ψi) |α|4 + |β|4 < 0
= −2(|α|4 + |β|4 − hψ|ν|ψi)
= −2((|α|2 + |β|2 )2 − 2|α|2 |β|2 − hψ|ν|ψi)
= −2(−2|α|2 |β|2 + 1 − hψ|ν|ψi)

where we used that hψ|ν|ψi is real. If ν is sufficiently close to ρ = |ψihψ|

then
1 − hψ|ν|ψi < 2|α|2 |β|2
and DB (ρ, σ(λ)) is increasing around σ. Thus we have found the minimum
4|α|2 (1 − |α|2 ).
(iv) For |α| ∈ [0, 1] we find

4|α|2 (1 − |α|2 ) ≤ −|α|2 log2 |α|2 − (1 − |α|2 ) log2 (1 − |α|2 ).

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 276

276 Problems and Solutions

Problem 19. The two-point Hubbard model with cyclic boundary condi-
tions is given by

Ĥ = t(c†1↑ c2↑ + c†1↓ c2↓ + c†2↑ c1↑ + c†2↓ c1↓ ) + U (n1↑ n1↓ + n2↑ n2↓ )

where
nj↑ := c†j↑ cj↑ , nj↓ := c†j↓ cj↓ , j = 1, 2.

The Fermi operators c†j↑ ,c†j↓ ,cj↑ ,cj↓ obey the anti-commutation relations

[c†j,σ , ck,σ0 ]+ = δσσ0 δjk I, [c†j,σ , c†k,σ0 ]+ = [cj,σ , ck,σ0 ]+ = 0.

Ĥ commutes with the total number operator N̂ , and the total spin operator
Ŝz in the z direction
2 2
1X †
(c†j↑ cj↑ + c†j↓ cj↓ ), (c cj↑ − c†j↓ cj↓ ).
X
N̂ := Ŝz :=
j=1
2 j=1 j↑

We consider the subspace with two electrons, N = 2 and Sz = 0. A basis

for 2 particles with total spin 0 is

|s1 i := c†1↑ c†1↓ |0i, |s2 i := c†1↑ c†2↓ |0i, |s3 i := c†2↑ c†1↓ |0i, |s4 i := c†2↑ c†2↓ |0i

where h0|0i = 1.
(i) Find the matrix representation of Ĥ in this basis.
(ii) Can the matrix representation of Ĥ be written in the form

Ĥ = A1 ⊗ I2 + I2 ⊗ A2

where A1 and A2 are 2 × 2 matrices and I2 is the 2 × 2 identity matrix?

Solution 19. (i) Applying Ĥ to the basis gives

Ĥ|s1 i = t|s2 i + t|s3 i + U |s1 i

Ĥ|s2 i = t|s1 i + t|s4 i
Ĥ|s3 i = t|s1 i + t|s4 i
Ĥ|s4 i = t|s2 i + t|s3 i + U |s4 i.

Identifying |si i with elements ei of the standard basis in C4 yields the

matrix representation of Ĥ

U t t 0
 
t 0 0 t
Ĥ =  .
t 0 0 t
0 t t U
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 277

Entanglement 277

(ii) Suppose a Hamilton operator K̂ can be written as K̂ = A1 ⊗I2 +I2 ⊗A2

where A1 , A2 ∈ M 2 and I2 is the 2 × 2 identity matrix. Then we have

exp(−iK̂τ /~) = exp(−iτ A1 /~ ⊗ I2 − iτ I2 /~ ⊗ A2 )

= exp(−iτ A1 /~) ⊗ exp(−iτ A2 /~).

In this case separable states remain separable under time evolution in the
model, and entangled states remain entangled under time evolution in the
model. For the matrix representation of Ĥ, however we have
 
0 1
Ĥ = tVN OT ⊗ I2 + tI2 ⊗ VN OT + diag(U, 0, 0, U ), VN OT := .
1 0

The diagonal matrix diag(U, 0, 0, U ) cannot be written in the form A1 ⊗

I2 +I2 ⊗A2 . Thus we conclude that almost all initial separable states evolve
into entangled states under the time evolution of the model.

Problem 20. Find the matrix representation of the two-point Hubbard

model in the basis

1 1
√ (c†1↓ c†1↑ |0i + c†2↓ c†2↑ |0i), √ (c†1↓ c†2↑ |0i + c†2↓ c†1↑ |0i),
2 2

1 1
√ (c†1↓ c†1↑ |0i − c†2↓ c†2↑ |0i), √ (c†1↓ c†2↑ |0i − c†2↓ c†1↑ |0i) .
2 2

Solution 20. The two-point Hubbard model admits a discrete symmetry

under the change 1 → 2, 2 → 1. Thus we have a finite group with two
elements. We obtain two irreducible representations. The group-theoretical
reduction leads to the two invariant subspaces
 
1 † † † † 1 † † † †
√ (c1↓ c1↑ |0i + c2↓ c2↑ |0i), √ (c1↓ c2↑ |0i + c2↓ c1↑ |0i) ,
2 2
 
1 1
√ (c†1↓ c†1↑ |0i − c†2↓ c†2↑ |0i), √ (c†1↓ c†2↑ |0i − c†2↓ c†1↑ |0i) .
2 2
These four states can be considered as the Bell states. In the Bell basis the
matrix representation of the Hubbard model is given by
U 2t 0 0
 
   
 2t 0 0 0  U 2t 0 0
= ⊕
0 0 0 0 2t 0 0 U

0 0 0 U
where ⊕ denotes the direct sum.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 278

278 Problems and Solutions

Problem 21. The two-point Hubbard model with cyclic boundary con-
ditions is given by
Ĥ = t(c†1↑ c2↑ + c†1↓ c2↓ + c†2↑ c1↑ + c†2↓ c1↓ ) + U (n1↑ n1↓ + n2↑ n2↓ ).
Find the time evolution of the initial state
1
|ψ(0)i = √ (c†1↑ c†1↓ − c†2↑ c†2↓ )|0i
2
under the two-point Hubbard model. When is the state |ψ(τ )i entangled?

Solution 21. Solving the Schrödinger equation

d
i~ |ψ(τ )i = Ĥ|ψ(τ )i
dτ
we find |ψ(τ )i = e−iU τ /~ |ψ(0)i. Consequently, the condition for separabil-
ity is given by  τ 
exp −2i U = 0.
~
This equation cannot be satisfied. Thus |ψ(τ )i is entangled for all τ .

Problem 22. An arbitrary pure state in the Hilbert space C4 can be

written as
cos(θ3 )
 
sin(θ3 ) cos(θ2 )eiφ3
|ψi = 
 
sin(θ3 ) sin(θ2 ) cos(θ1 )eiφ2

sin(θ3 ) sin(θ2 ) sin(θ1 )eiφ1
where θk ∈ [0, π/2], and φk ∈ [0, 2π) for k = 1, 2, 3.
(i) Find values for θk and φk (k = 1, 2, 3) so that we obtain the unentangled
state (product state)
1
 
   
0 1 1
|αi =   ≡ ⊗ .
0 0 0
0
(ii) Find values for θk and φk (k = 1, 2, 3) so that we obtain the entangled
(Bell) state
1
 
1 0
|βi = √   .
2 0
1

Solution 22. (i) We obtain the unentangled state by setting θ3 = 0,

θ2 = 0 and θ1 = 0.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 279

Entanglement 279

(ii) We obtain the Bell state by setting θ3 = π/4, θ2 = π/2, θ1 = π/2 and
φ3 = φ2 = φ1 = 0.

Problem 23. A completely entangled state |ΨAB i of an (N × N )-

dimensional Hilbert space H ⊗ H can be written as
N −1
1 X A
|ΨAB i = √ |ψk i ⊗ |ψkB i
N k=0

where { |ψk i : k = 0, 1, . . . , N − 1 } is a orthonormal basis of the Hilbert

space H. We define the linear operators (unitary depolarizers)
N
X −1
Ujk := e(2πi/N )j` |ψ` mod N ihψ`+k mod N |
`=0

where j, k = 0, 1, . . . , N − 1.
(i) Calculate
N −1 N −1
1 X X † 1 †
Ujk XUjk , tr(Ujk U`m )
N j=0 N
k=0

where X is an arbitrary linear operator defined on the Hilbert space H.

(ii) Discuss the set

{ |ΨAB A
jk i = (Ujk ⊗ IB )|Ψ
AB
i : j, k = 0, 1, . . . , N − 1 }.

Solution 23. (i) We obtain

N −1 N −1
1 X X †
Ujk XUjk = (tr(X))I
N j=0
k=0

where I is the identity operator in H. We obtain

1 †
tr(Ujk U`m ) = δj` δkm .
N
(ii) The set is an orthonormal basis of the Hilbert space H ⊗ H.

Problem 24. Consider the GHZ state (Greenberger-Horne-Zeilinger state)

1
|GHZi = √ (|0i ⊗ |0i ⊗ |0i + |1i ⊗ |1i ⊗ |1i)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 280

280 Problems and Solutions

and the W state

1
|W i = √ (|0i ⊗ |0i ⊗ |1i + |0i ⊗ |1i ⊗ |0i + |1i ⊗ |0i ⊗ |0i).
3
(i) Calculate the states
√ √
( 2h0| ⊗ I2 ⊗ I2 )|GHZi, ( 2I2 ⊗ h0| ⊗ I2 )|GHZi
√ √
( 2I2 ⊗ I2 ⊗ h0|)|GHZi, ( 2h1| ⊗ I2 ⊗ I2 )|GHZi
√ √
( 2I2 ⊗ h1| ⊗ I2 )|GHZi, ( 2I2 ⊗ I2 ⊗ h1|)|GHZi
and discuss.
(iii) Calculate
√ ! √ !
3 3
√ h0| ⊗ I2 ⊗ I2 |W i, √ I2 ⊗ h0| ⊗ I2 |W i,
2 2
√ !
3 √
√ I2 ⊗ I2 ⊗ h0| |W i, ( 3h1| ⊗ I2 ⊗ I2 )|W i,
2
√ √
( 3I2 ⊗ h1| ⊗ I2 )|W i, ( 3I2 ⊗ I2 ⊗ h1|)|W i
and discuss.

Solution 24. (i) We find

√ √
( 2h0| ⊗ I2 ⊗ I2 )|GHZi = |0i ⊗ |0i, ( 2I2 ⊗ h0| ⊗ I2 )|GHZi = |0i ⊗ |0i
√ √
( 2I2 ⊗ I2 ⊗ h0||GHZi = |0i ⊗ |0i, ( 2h1| ⊗ I2 ⊗ I2 )|GHZi = |1i ⊗ |1i
√ √
( 2I2 ⊗ h1| ⊗ I2 )|GHZi = |1i ⊗ |1i, ( 2I2 ⊗ I2 ⊗ h1|)|GHZi = |1i ⊗ |1i.
Thus all the two-particle states are not entangled after measurement of the
third state, although the |GHZi state is entangled. The |GHZi state is
usually referred to as maximally entangled in several senses, e.g. it violates
Bell inequalities maximally. However, from the result above we find that
the state is maximally fragile, i.e. if one particle is lost or projected onto
the computational basis { |0i, |1i }, then all entanglement is destroyed.
(ii) We find
√ !
3 1
√ h0| ⊗ I2 ⊗ I2 |W i = √ (|0i ⊗ |1i + |1i ⊗ |0i)
2 2
√ !
3 1
√ I2 ⊗ h0| ⊗ I2 |W i = √ (|0i ⊗ |1i + |1i ⊗ |0i)
2 2
√ !
3 1
√ I2 ⊗ I2 ⊗ h0| |W i = √ (|0i ⊗ |1i + |1i ⊗ |0i).
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 281

Entanglement 281

Thus in this case the projected states are (maximally) entangled (Bell
states). However, for the other cases we find
√
( 3h1| ⊗ I2 ⊗ I2 )|W i = |0i ⊗ |0i
√
( 3I2 ⊗ h1| ⊗ I2 )|W i = |0i ⊗ |0i
√
( 3I2 ⊗ I2 ⊗ h1|)|W i = |0i ⊗ |0i.
Thus these states are not entangled.

Problem 25. Let σ1 , σ2 and σ3 be the Pauli spin matrices. Consider the
Hamilton operator
1
(σ3 ⊗ I2 + I2 ⊗ σ3 ) − ∆(σ1 ⊗ σ1 )
Ĥ =
2
where  > 0 and ∆ > 0. Find the eigenvalues and normalized eigenvectors
of Ĥ. Are the eigenvectors entangled?

Solution 25. Let

   
1 0
| ↑i = , | ↓i =
0 1
and | ↑↑i = | ↑i⊗| ↑i etc.. Then the eigenvalues and normalized eigenvectors
are given by
p 1
E0 = − 2 + ∆2 , |0i = √ (| ↓↓i + a| ↑↑i)
1 + a2
1
E1 = −∆, |1i = √ (| ↑↓i + | ↓↑i)
2
1
E2 = +∆, |2i = √ (−| ↑↓i + | ↓↑i)
2
p 1
2
E3 = +  + ∆ , 2 |3i = √ (−a| ↓↓i + | ↑↑i)
1 + a2
where √
2 + ∆ 2 − 
a := .
∆
The eigenstates are entangled.

Problem 26. Consider the Hilbert space H = Cn and the product space
H ⊗ H. Let A be an arbitrary n × n matrix over C and In the n × n identity
matrix. Consider the following definition. A normalized vector ψ ∈ H ⊗ H
is called maximally entangled, if its reduced density matrix is maximally
mixed, i.e. a multiple of In
hψ|(A ⊗ In )|ψi = dim(H)−1 tr(A). (1)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 282

282 Problems and Solutions

(i) Show that the Bell states in C4

1 1 0 0
       
1 0 1  0  1 1 1  1 
Φ+ = √   , Φ− = √  + −
, Ψ = √  , Ψ = √ 
2 0 0 2 1 2 −1

2
1 −1 0 0
satisfy equation (1), where H = C2 .
(ii) Calculate the left and right-hand side of equation (1) for the vector
(1 0 0 0). Discuss.

Solution 26. (i) We have dim(H) = n = 2 and tr(A) = a11 + a22 . For
the left-hand side we have
a11 0 a12 0
 
 0 a11 0 a12 
(A ⊗ In ) =  .
a21 0 a22 0
0 a21 0 a22
Thus for the Bell state Ψ+ we have
a11 0 a12 0 1
  
1  0 a11 0 a12   0  1
(1 0 0 1)    = (a11 + a22 ).
2 a21 0 a22 0 0 2
0 a21 0 a22 1
Analogously we prove for the other Bell states that equation (1) is satisfied.
(ii) For the vector (1 0 0 0) we obtain for the left hand side a11 . Thus
equation (1) is not satisfied. This state is not entangled.

Problem 27. Let |Hi (|V i) indicate the state of a horizontal (vertical)
polarized photon. Suppose we have the product state
1 1
|ψi1234 = √ (|Hi1 ⊗ |V i2 − |V i1 ⊗ |Hi2 ) ⊗ √ (|Hi3 ⊗ |V i4 − |V i3 ⊗ |Hi4 ).
2 2
Thus we have a product state of two polarization entangled pairs. One
photon out of each pair (2 and 3) is directed to the two inputs of a polarizing
beam splitter. What is the output?

Solution 27. Since the polarizing beam splitter transmits horizontally

polarized photons and reflects vertically polarized photons, coincidence de-
tection between the two polarizing beam splitter outputs implies that either
both photons 2 and 3 are both horizontally polarized or both vertically po-
larized. Thus the state (1) is projected onto a two-dimensional subspace
spanned by
|V i1 ⊗ |Hi2 ⊗ |Hi3 ⊗ |V i4 , |Hi1 ⊗ |V i2 ⊗ |V i3 ⊗ |Hi4 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 283

Entanglement 283

After the polarizing beam splitter, the renormalized state corresponding to

a fourfold coincidence is
1
|φi120 30 4 = √ (|Hi1 ⊗ |V i20 ⊗ |V i30 ⊗ |Hi4 + |V i1 ⊗ |Hi20 ⊗ |Hi30 ⊗ |V i4 ).
2
This is a GHZ state of four particles, which can exhibit nonlocal behaviour
according to the GHZ theorem.

Problem 28. An entanglement witness W on the product Hilbert space

HA ⊗ HB is a linear operator on HA ⊗ HB such that W is not positive
semi-definite and
hψ|W |ψi ≥ 0
for all separable |ψi = |ψA i ⊗ |ψB i, where |ψA i ∈ HA and |ψB i ∈ HB .
Show that
0 0 0 −1
 
0 1 0 0 
W = I4 − 2|Φ+ ihΦ+ | = 

0 0 1 0

−1 0 0 0
is an entanglement witness.

Solution 28. The eigenvalues of W are −1 and 1 with multiplicity 3,

with corresponding eigenvectors given by the Bell basis. Clearly W is not
positive semi-definite. Using |ψi = |ψA i ⊗ |ψB i, with |ψA i, |ψB i ∈ C2 , we
find
2
hψ|W |ψi = hψ
 A |ψA ihψB |ψB i − |hψ
A |0ihψB |0i + hψA |1ihψ
B |1i|
= |hψA |0i|2 + |hψA |1i|2 |hψB |0i|2 + |hψB |1i|2
2
−
 |hψA |0ihψB |0i + hψA
|1ihψB |1i| 
2 2
≥ |hψA |0i| + |hψA |1i| |hψB |0i|2 + |hψB |1i|2
 2
− |hψA |0ihψB |0i| + |hψA |1ihψB |1i|
 2
= |hψA |0ihψB |1i| − |hψA |1ihψB |0i| ≥ 0.

Thus the operator W is an entanglement witness. Furthermore,

tr(W |Φ+ ihΦ+ |) = hΦ+ |W |Φ+ i = −1.

Consequently W is a witness for the entangled Bell state |Φ+ i.

Problem 29. Consider the distillation under 1-local operations of a pure

state in the Hilbert space C4

|ψi := α|00i + β|11i

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 284

284 Problems and Solutions

with |α|2 + |β|2 = 1. If this state can be distilled, it will be transformed

into the Bell state
1 1
|Φ+ i = √ |00i + √ |11i.
2 2
Find a generalized measurement on the first qubit which can be used to
distill |Φ+ i from |ψi.

Solution 29. We could attempt to construct a separable unitary transfor-

mation to perform this transform, however one component of a generalized
measurement on the first qubit will achieve this for k ∈ C, k 6= 0
k k
U0 := |0ih0| + |1ih1|
α β
which transforms the density operator ρ := |ψihψ| to
(U0 ⊗ I)ρ(U0 ⊗ I)∗
= |Φ+ ihΦ+ |
tr(U0 ⊗ I)ρ(U0 ⊗ I)∗
with probability

p0 := tr((U0 ⊗ I)ρ(U0 ⊗ I)∗ ) = 2|k|2 .

Of course this is only possible for |α2 |, |β|2 6= 0, i.e. for α = 0 or β = 0 there
is no distillable entanglement. To complete the generalized measurement
we must find U1 such that U0∗ U0 + U1∗ U1 = I. Since U0∗ U0 is clearly positive
semidefinite we can use the polar decomposition U1 = U1 H1 , where U1 is
unitary and H1 is positive semidefinite. We obtain
s s 
 2  2
p k k
U1 = U1 I − U0∗ U0 = U1  1 −   |0ih0| + 1 −   |1ih1| .
α β

Consequently it is necessary that k ≤ min{|α|2 , |β|2 }.

Problem 30. Consider the GHZ state (Greenberger-Horne-Zeilinger)

1
|ψi := √ (|001i + |110i).
2
We consider the basis α = {|Li, |Ri} and the basis β = {|Hi, |V i} described
by
1 1
α: |0i = √ (|Li + |Ri), |1i = √ (|Li − |Ri),
2 2
1 i
β: |0i = √ (|Hi + |V i), |1i = √ (|Hi − |V i).
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 285

Entanglement 285

(i) Express |ψi in terms of the bases α and β, where one qubit is in the
basis α and the remaining two in the basis β.
(ii) Use (i) to predict the measurement outcomes where all qubits are mea-
sured in the basis α. Compare with the actual outcomes.

Solution 30. (i) Expressing only one qubit of |ψi in the basis α and the
rest in the basis β basis yields

|ψi = 2i (|LHHi + |RV Hi − |LV V i − |RHV i) αββ

i
= 2 (|HLHi + |V RHi − |V LV i − |HRV i) βαβ
1
= 2 (|HV Li + |V HLi − |HHRi − |V V Ri) ββα

Measuring two qubits in the β basis as given in the previous equations

allows us to deduce the result of measuring the other qubit in the α basis.
Let βj denote the result after measuring qubit j in the β basis. As an
example, from the first equality, if β2 = β3 then the first qubit is |Li, and
|Ri when β2 6= β3 . Thus we construct the following table.

Outcomes in β basis Outcomes in α basis

β1 = β2 = β3 LLR
β1 = β2 6= β3 RRR
β1 = β3 6= β2 RLL
β2 = β3 6= β1 LRL

(ii) We find in the α basis

|ψi = 12 (|LLLi + |RRLi − |LRRi − |RLRi).

None of the results obtained are consistent with the outcomes in the table
deduced in (i).

Problem 31. Consider the state in C8

1
X
|ψi = cj1 j2 j3 |j1 i ⊗ |j2 i ⊗ |j3 i
j1 ,j2 ,j3 =0

where    
1 0
|0i = , |1i = .
0 1
Let R, S, T be 2 × 2 matrices over C
     
r11 r12 s11 s12 t11 t12
R= , S= , T =
r21 r22 s21 s22 t21 t22
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 286

286 Problems and Solutions

where det(R) = 1, det(S) = 1, det(T ) = 1. This means that R, S and T

are elements of the Lie group SL(2, C). Let
1
X 1
X
dj1 j2 j3 |j1 i⊗|j2 i⊗|j3 i = (R⊗S ⊗T ) cj1 j2 j3 |j1 i⊗|j2 i⊗|j3 i.
j1 ,j2 ,j3 =0 j1 ,j2 ,j3 =0
(1)
We define the hyperdeterminant of C = (cj1 j2 j3 ) with j1 , j2 , j3 ∈ { 0, 1 } as
Det(C) := (c2000 c2111 + c2001 c2110 + c2010 c2101 + c2011 c2100 )
−2(c000 c001 c110 c111 + c000 c010 c101 c111 + c000 c011 c100 c111
+c001 c010 c101 c110 + c001 c011 c110 c100 + c010 c011 c101 c100 )
+4(c000 c011 c101 c110 + c001 c010 c100 c111 ).
Show that
Det(C) = Det(D). (2)
Owing to (2) the quantity Det(C) is called an invariant.

Solution 31. Note that det(R ⊗ S ⊗ T ) = 1. We find eight equations for

d000 , d001 , . . . , d111 . From det(R) = det(S) = det(T ) = 1 it follows that
r11 r22 − r12 r21 = 1, s11 s22 − s12 s21 = 1, t11 t22 − t12 t21 = 1. Inserting dj1 j2 j3
into the right-hand side of (2) and inserting these conditions it follows that
Det(D) = Det(C).

Problem 32. The hyperdeterminant of a 2×2×2 hypermatrix C = (cijk )

(i, j, k ∈ { 0, 1 }) is defined by
1 1
1 X X
Det(C) := − ii0 jj 0 kk0 mm0 nn0 pp0 cijk ci0 j 0 m cnpk0 cn0 p0 m0
2
i,j,k,m,n,p=0 i0 ,j 0 ,k0 ,m0 ,n0 ,p0 =0

where 00 = 11 = 0, 01 = 1, 10 = −1.

(i) Calculate Det(C).
(ii) Consider the three qubit state
1
X
|ψi = cijk |ii ⊗ |ji ⊗ |ki.
i,j,k=0

The three tangle τ3 is a measure of entanglement and is defined for the three
qubit state |ψi as
τ123 := 4|Det(C)|
where C = (cijk ). Find the three tangle for the GHZ state
1
|GHZi = √ (|0i ⊗ |0i ⊗ |0i + |1i ⊗ |1i ⊗ |1i)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 287

Entanglement 287

and the W state

1
|W i = √ (|0i ⊗ |0i ⊗ |1i + |0i ⊗ |1i ⊗ |0i + |1i ⊗ |0i ⊗ |0i).
3

Solution 32. (i) We obtain

Det(C) := (c2000 c2111 + c2001 c2110 + c2010 c2101 + c2011 c2100 )
−2(c000 c001 c110 c111 + c000 c010 c101 c111 + c000 c011 c100 c111
+c001 c010 c101 c110 + c001 c011 c110 c100 + c010 c011 c101 c100 )
+4(c000 c011 c101 c110 + c001 c010 c100 c111 ).
√ √
(ii) For the GHZ state we obtain c000 = 1/ 2, c111 = 1/ 2. All other
coefficients are zero. Thus we find for the three tangle τ3 = 1. For the W
state we have c001 = c010 = c100 = √13 . Thus τ3 = 0. Using this measure of
entanglement the W state is not entangled. Note that the W -state cannot
be written as a product state.

Problem 33. Let |0i, |1i be an orthonormal basis in C2 . Consider the

normalized state
X1
|ψi = cjk |ji ⊗ |ki
j,k=0
4
in the Hilbert space C and the 2 × 2 matrix C = (cjk ). Using the 4
coefficients cjk (j, k ∈ {0, 1}) we form a multilinear polynomial p in two
variables x1 , x2
p(x1 , x2 ) = c00 + c01 x1 + c10 x2 + c11 x1 x2 . (1)
Show that determinant det(C) = c00 c11 − c01 c10 is the unique irreducible
polynomial (up to sign) of content one in the 4 unknowns cjk that vanishes
whenever the system of equations
∂p ∂p
p= = =0 (2)
∂x1 ∂x2
has a solution (x∗1 , x∗2 ) in C2 .

Solution 33. Inserting (2) into (1) provides the three equations
c00 + c01 x1 + c10 x2 + c11 x1 x2 = 0, c01 + c11 x2 = 0, c10 + c11 x1 = 0.
Multiplication of the first equation with c11 and inserting the second and
third equation yields
c00 c11 + c01 c11 x1 + c10 c11 x2 + c11 x1 c11 x2 = c00 c11 − c01 c10 − c10 c01 + c10 c10
= c00 c11 − c01 c10
= 0.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 288

288 Problems and Solutions

Problem 34. Consider the state

|ψi = cos(α)|00i + sin(α)|11i, 0 < α < π/4
where α is called the Schmidt angle.
(i) Find the eigenvalues of the density matrix |ψihψ|.
(ii) Find the partially traced density matrix (we find when we trace over
one of the subsystems).
(iii) Show that the partially traced density matrix has two unequal and
non-zero eigenvalues λ1 = cos2 (α) and λ2 = sin2 (α).
(iv) Calculate the von Neumann entropy for the corresponding density ma-
trix. Show that the entropy grows monotonically with the Schmidt angle.

Solution 34. (i) We obtain for the density matrix

cos2 (α) 0 0 sin(2α)/2
 
0 0 0 0
ρ(α) = 
 
0 0 0 0

2
sin(2α)/2 0 0 sin (α)
with the eigenvalues 0 (triple) and 1. Thus the eigenvalues are independent
of α.
(ii) We obtain the diagonal matrix
 
cos2 (α) 0
ρ2 = .
0 sin2 (α)

(iii) From (ii) we see that the eigenvalues are cos2 (α) and sin2 (α).
(iv) Thus for the entropy we find
S(ρ2 ) = − cos2 (α) log2 (cos2 (α)) − sin2 (α) log2 (sin2 (α))
which is monotonically increasing for 0 < cos2 (α) < 1/2, i.e. 0 < α < π/4.

Problem 35. Let σ1 , σ2 , σ3 be the Pauli spin matrices. We form the

nine 4 × 4 matrices
Σjk := σj ⊗ σk , j, k = 1, 2, 3.

Note that [Σjk , Σmn ] = 04 . The variance of an hermitian operator Ô and

a wave vector |φi is defined by
VÔ (|φi) := hφ|(Ô)2 |φi − (hφ|Ô|φi)2 .
The remoteness for a given normalized state |ψi in C4 is defined by
3 X
X 3
hψ|(Σjk )2 |ψi − (hψ|Σjk |ψi)2 .


R(|ψi) =
j=1 k=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 289

Entanglement 289

Find the remoteness for the Bell states

1 1
|φ+ i = √ (|0i ⊗ |0i + |1i ⊗ |1i), |φ− i = √ (|0i ⊗ |0i − |1i ⊗ |1i)
2 2

1 1
|ψ + i = √ (|0i ⊗ |1i + |1i ⊗ |0i), |ψ − i = √ (|0i ⊗ |1i − |1i ⊗ |0i).
2 2

Solution 35. We have for all j, k = 1, 2, 3

Σ2jk = I2 ⊗ I2 = I4 .

Thus hψ|(Σjk )2 |ψi = 1, where |ψi is one of the Bell states. Now

hψ|(σ1 ⊗ σ1 )|ψi = hψ|(σ2 ⊗ σ2 )|ψi = hψ|(σ3 ⊗ σ3 )|ψi = 1.

All the other matrices of Σjk yield 0 for all Bell states. Thus we find for
the remoteness of all Bell states R = 6.

Problem 36. A general pure state |Ψi of two qubits can be written as

|Ψi = eiφ0 cos θ0 |00i + eiφ1 sin(θ0 ) cos(θ1 )|01i

+eiφ2 sin(θ0 ) sin(θ1 ) cos(θ2 )|10i + eiφ3 sin(θ0 ) sin(θ1 ) sin(θ2 )|11i(1)

where φj and θk are chosen uniformly according to the Haar measure

1
dµ = d(sin(θ0 ))6 d(sin(θ1 ))4 d(sin(θ2 ))2 dφ0 dφ1 dφ2 dφ3 (2)
(2π)4

with
π
0 ≤ φj < 2π, 0 ≤ θk < (3)
2
where j = 0, 1, 2, 3 and k = 0, 1, 2. An extra overall random phase eiφ0
is included to maintain consistency with SU (n), where n = 4. For a pure
state of two qubits the tangle τ , is defined as

τ := 4 det(ρA ) (4)

where ρA is the reduced density matrix obtained when qubit B has been
traced over (or vice versa, permuting A and B). The tangle τ is an entan-
glement measure.
(i) Find τ for |ψi. Then find τ for the four Bell states and the unentangled
state |00i.
(ii) Using the Haar measure find τ for a randomly selected pure state.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 290

290 Problems and Solutions

Solution 36. (i) From the state |Ψi we obtain the 4 × 4 density matrix

ψ0 ψ0∗ ψ0 ψ1∗ ψ0 ψ2∗ ψ0 ψ3∗

 
∗ ∗ ∗ ∗
 ψ ψ ψ1 ψ1 ψ1 ψ2 ψ1 ψ3 
ρ = |ΨihΨ| =  1 0∗ (5)
ψ2 ψ0 ψ2 ψ1∗ ψ2 ψ2∗ ψ2 ψ3∗

∗ ∗ ∗ ∗
ψ3 ψ0 ψ3 ψ1 ψ3 ψ2 ψ3 ψ3
where
ψ0 = eiφ0 cos(θ0 ), ψ1 = eiφ1 sin(θ0 ) cos(θ1 )
ψ2 = eiφ2 sin(θ0 ) sin(θ1 ) cos(θ2 ), ψ3 = eiφ3 sin(θ0 ) sin(θ1 ) sin(θ2 ). (6)
Using the basis    
1 0
⊗ I2 , ⊗ I2 (7)
0 1
where I2 is the 2 × 2 unit matrix we find the 2 × 2 matrix

ψ0 ψ0∗ + ψ2 ψ2∗ ψ0 ψ1∗ + ψ2 ψ3∗

 
ρA = . (8)
ψ1 ψ0∗ + ψ3 ψ2∗ ψ1 ψ1∗ + ψ3 ψ3∗

It follows that

det(ρA ) = (ψ0 ψ0∗ +ψ2 ψ2∗ )(ψ1 ψ1∗ +ψ3 ψ3∗ )−(ψ1 ψ0∗ +ψ3 ψ2∗ )(ψ0 ψ1∗ +ψ2 ψ3∗ ). (9)

Therefore

det(ρA ) = ψ0 ψ0∗ ψ3 ψ3∗ + ψ1 ψ1∗ ψ2 ψ2∗ − ψ0 ψ1∗ ψ2∗ ψ3 − ψ0∗ ψ1 ψ2 ψ3∗ . (10)

Inserting (6) into (10) we get

det(ρA ) = cos2 (θ0 ) sin2 (θ0 ) sin2 (θ1 ) sin2 (θ2 ) + sin4 (θ0 ) cos2 (θ1 ) sin2 (θ1 ) cos2 (θ2 )
−(ei(φ0 −φ1 −φ2 −φ3 ) + ei(−φ0 +φ1 +φ2 −φ3 ) )
× sin3 (θ0 ) cos(θ0 ) sin2 (θ1 ) cos(θ1 ) sin(θ2 ) cos(θ2 ).

It follows that the four Bell states have the maximum possible entangle-
ment, i.e. τ = 1. The product state |00i has τ = 0.
(ii) From (2) we find
48
cos(θ0 )(sin(θ0 ))5 cos(θ1 )(sin θ1 )3 cos(θ2 ) sin(θ2 )dθ0 dθ1 dθ2 dφ0 dφ1 dφ2 dφ3
(2π)4
and Z
dµ = 1
SU (4)

i.e. the Haar measure is normalized. Here we made use of

Z 2π Z π/2
1
dφ = 2π, and sink (x) cos(x)dx =
0 0 k+1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 291

Entanglement 291

where k = 1, 2, . . .. Integrating det(ρA ) (or det(ρB )) over the Haar measure

gives hτ i = 25 , where we used
π/2 π/2
m−1
Z Z
m n
sin (x) cos (x)dx = sinm−2 (x) cosn (x)dx
0 m+n 0

π/2 π/2
n−1
Z Z
sinm (x) cosn (x)dx = sinm (x) cosn−2 (x)dx
0 m+n 0
and Z π/2
1
sin(x) cos(x)dx = .
0 2
A randomly selected pure state of two qubits might thus be expected to
have 0.4 tangle units of entanglement.

Programming Problems

Problem 1. Let |ψi be a given state in the Hilbert space Cn . Let X and
Y be two n × n hermitian matrices. We define the correlation for a given
state |ψi as

CXY (|ψi) := hψ|XY |ψi − hψ|X|ψihψ|Y |ψi.

Let n = 4,

0 0 0 1 1 0 0 0
   
0 0 1 0 0 0 0 1
X= , Y =
0 1 0 0 0 0 1 0

1 0 0 0 0 1 0 0

and consider the Bell state

1
 
1 0
|ψi = √   .
2 0
1

Find the correlation.

Solution 1. Since X|ψi = |ψi we have hψ|X|ψi = 1 and hψ|X = hψ|.

Thus

hψ|XY |ψi − hψ|X|ψihψ|Y |ψi = hψ|Y |ψi − hψ|Y |ψi = 0.

A Maxima implementation is
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 292

292 Problems and Solutions

/* CorrelationBell.mac */
X: matrix([0,0,0,1],[0,0,1,0],[0,1,0,0],[1,0,0,0]);
Y: matrix([1,0,0,0],[0,0,0,1],[0,0,1,0],[0,1,0,0]);
b: matrix([1/sqrt(2)],[0],[0],[1/sqrt(2)]);
bT: transpose(b);
CXYb: bT . X . Y . b - (bT . X . b) . (bT . Y . b);

Problem 2. Consider the normalized vector in C6

1
 
 −1 
1 0 

v=  .
2 0 
1
 
−1
Let u be a vector in C3 and w be a vector in C2 . Can v be written as
v = u ⊗ w? Can v be written as v = w ⊗ u?

Solution 2. In the first case we have to solve the system of equations

1/2 u1 w1
   
 −1/2   u1 w2 
 0   u2 w1 
   
= .
 0   u2 w2 

1/2 u3 w1
   
−1/2 u3 w2
In the second case we have to solve the system of equations
1/2 w1 u1
   
 −1/2   w1 u2 
 0   w1 u3 
   
= .
 0   w2 u1 

1/2 w2 u2
   
−1/2 w2 u3
For the first system the Maxima program
/* C6.mac */
solve([1/2-u1*w1=0,-1/2-u1*w2=0,u2*w1=0,u2*w2=0,1/2-u3*w1=0,-1/2-u3*w2],
[u1,u2,u3,w1,w2]);
solve([1/2-w1*u1=0,-1/2-w1*u2=0,w1*u3=0,w2*u1=0,1/2-w2*u2=0,-1/2-w2*u3],
[u1,u2,u3,w1,w2]);

provides the solution (r arbitrary and r 6= 0)

1 1
u1 = , u2 = 0, u3 = , w1 = r, w2 = −r.
2r 2r
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 293

Entanglement 293

Hence we can write

1
 
 −1   
1  
1 0
 1 1 1
= √ 0 ⊗√ .
   
2 0  2 −1

2 1
1
 
−1
The solution set for the second system is empty, i.e. v cannot be written
as v = w ⊗ u.

10.3 Supplementary Problems

Problem 1. Consider the two Hilbert spaces H1 = H2 = Cd and the

product Hilbert space H = H1 ⊗ H2 . A state |ψi ∈ H is called maximally
entangled if
1
trH1 (|ψihψ|) = trH2 (|ψihψ|) = .
d
Apply this definition to the Bell states in H = C4 , i.e. d = 2
1 1
   
1 0 1  0 
|ψ1 i = √   , |ψ2 i = √  ,
2 0 2 0
1 −1
0 0
   
1 1 1  1 
|ψ3 i = √   , |ψ4 i = √  .
2 1 2 −1
0 0

Problem 2. (i) The normalized states

−1 1 1 1
       
1  1  1  −1  1  1  1  1 
, , ,
2 1 2 1 2 −1 2 1
    
1 1 1 −1
form an orthonormal basis in C4 . Are the states entangled?
(ii) Do the normalized states
−1 0 1 1
       
1  1  1  −1  1  0  1  1 
√  , √  , √  , √ 
1 1 3 −1 0

3 3 3
0 1 1 −1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 294

294 Problems and Solutions

form an orthonormal basis in C4 ? Are the states entangled?

Problem 3. Show that the GHZ state

1 T
|ψi = √ ( 1 0 0 0 0 0 0 1)
2

in C8 is entangled.

n
Problem 4. We consider the finite-dimensional Hilbert space H = C2
and the normalized state
1
X
|ψi = cj1 ,j2 ,...,jn |j1 i ⊗ |j2 i ⊗ · · · ⊗ |jn i
j1 ,j2 ,...,jn =0

in this Hilbert space. Here |0i, |1i denotes the standard basis. Let jk
(j, k = 0, 1) be defined by 00 = 11 = 0, 01 = 1, 10 = −1. Let n be even
or n = 3. Then an n-tangle can be introduced by



 X 1
τ1...n = 2  cα1 ...αn cβ1 ...βn cγ1 ...γn cδ1 ...δn


 α1 ,...,α
...
n =0
 δ ,...,δ
1 n =0

×α1 β1 α2 β2 · · · αn−1 βn−1 γ1 δ1 γ2 δ2 · · · γn−1 δn−1 αn γn βn δn  .
√ √
Consider the case n = 4 and a state |ψi with c0000 = 1/ 2, c1111 = 1/ 2
and all other coefficients are 0. Show that τ1234 = 1.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 295

Chapter 11

Bell Inequality

11.1 Introduction
Bell’s theorem states that, according to quantum mechanics, the value of
a certain combination of correlations for experiments of two distant sys-
tems can be higher than the highest value allowed by any local-realistic
theory of the type proposed by Einstein, Podolsky and Rosen, in which
local properties of a system determine the result of any experiment on that
system. The most discussed Bell inequality, the Clauser-Horne-Shimony-
Holt (CHSH) inequality states that in any local-realistic theory the absolute
value of a combination of four correlations is bounded by 2.

Consider a system with two distant particles i and j. Let A and a (B

and b) be physical observables taking values −1 and 1 referring to local
experiments on particle i (j). The correlation C(A, B) of A and B is defined
as

C(A, B) := PAB (1, 1) − PAB (1, −1) − PAB (−1, 1) + PAB (−1, −1)

where PAB (1, −1) denotes the joint probability of obtaining A = 1 and
B = −1 when A and B are measured. In any theory in which local variables
of particle i (j) determine the result of local experiments on particle i (j),
the absolute value of a particular combination of correlations is bounded
by 2
|C(A, B) − mC(A, b) − nC(a, B) − mnC(a, b)| ≤ 2

295
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 296

296 Problems and Solutions

where m and n can either be −1 or 1. The CHSH inequality holds for any
local-realistic theory, where m, n ∈ { −1, 1 }. For a two particle system in
a quantum pure state |ψi, the quantum correlation of A and B is defined

CQ (A, B) := hψ|ÂB̂|ψi

where  and B̂ are the self-adjoint operators which represent the observ-
ables A and B. For certain choices of Â, â, B̂, b̂, and |ψi, the quantum
correlation violates the CHSH inequality.

Let HA and HB be finite-dimensional Hilbert spaces. Let H be the Hilbert

space H = HA ⊗ HB , i.e. H is the tensor product of the two Hilbert spaces
HA and HB . Let Â1 , Â2 be hermitian operators (matrices) in HA with

Â21 = IA , Â22 = IA

and let B̂1 , B̂2 be hermitian operators (matrices) in HB with

B̂12 = IB , B̂22 = IB .

Let |ψi be a normalized state in the product Hilbert space HA ⊗ HB . The

generalized Bell inequality is given by

|hψ|Â1 ⊗ B̂1 |ψi + hψ|Â1 ⊗ B̂2 |ψi + hψ|Â2 ⊗ B̂1 |ψi − hψ|Â2 ⊗ B̂2 |ψi| ≤ 2.

If |ψi can be written as a product state |ψi = |φA i ⊗ |φB i then the Bell
inequality is not violated.

The Pauli spin matrices σ1 , σ2 , σ3 admit the eigenvalues +1 and −1 and

the 2 × 2 identity matrix admits the eigenvalue +1. Hence all Kronecker
products of these matrices admits these eigenvalues.
Let
A1 = σ3 ⊗ I2 , A2 = σ1 ⊗ I2 , A3 = σ2 ⊗ I2
1 1
B1 = I2 ⊗ √ (σ3 + σ1 − σ2 ), B2 = I2 ⊗ √ (σ3 − σ1 + σ2 )
3 3
1 1
B3 = I2 ⊗ √ (−σ3 + σ1 + σ2 ), B4 = I2 ⊗ √ (−σ3 − σ1 − σ2 )
3 3
Σ := A1 (B1 +B2 −B3 −B4 )+A2 (B1 −B2 +B3 −B4 +A3 (B1 −B2 −B3 +B4 )
and |ψi = √1 (|0iA ⊗ |0iB + |1iA ⊗ |1iB ). Then
2
√
hψ|Σ|ψi = 4 3.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 297

Bell Inequality 297

11.2 Solved Problems

Problem 1. Let HA and HB be finite-dimensional Hilbert spaces. Let
H be the Hilbert space H = HA ⊗ HB , i.e. H is the tensor product of
the two Hilbert spaces HA and HB . Let |ψi be a normalized vector (pure
state) in H. Let X be an observable (described as a hermitian matrix X̂)
in H. Then hψ|X̂|ψi defines the expectation values. The following three
conditions are equivalent when applied to pure states.

1. Factorisability: |ψi = |αi ⊗ |βi, where |αi ∈ HA and |βi ∈ HB with |αi
and |βi normalized.

2. The generalized Bell inequality: Let Â1 , Â2 be hermitian operators

(matrices) in HA with

Â21 = IA , Â22 = IA

where IA is the identity operator in HA . Let B̂1 , B̂2 be hermitian operators

(matrices) in HB with

B̂12 = IB , B̂22 = IB

where IB is the identity operator in HB . Thus the eigenvalues of Â1 , Â2 ,

B̂1 and B̂2 can only be ±1. The generalized Bell inequality is

|hψ|Â1 ⊗ B̂1 |ψi + hψ|Â1 ⊗ B̂2 |ψi + hψ|Â2 ⊗ B̂1 |ψi − hψ|Â2 ⊗ B̂2 |ψi| ≤ 2.

3. Statistical independence: For all hermitian operators  on HA and B̂ on

HB with the conditions given above

hψ|Â ⊗ B̂|ψi = hψ|Â ⊗ IB |ψihψ|IA ⊗ B̂|ψi.

(i) Show that condition 3 follows from condition 1.

(ii) Show that condition 2 follows from condition 3.

Solution 1. (i) Consider the product state |ψi = |αi ⊗ |βi. Then

hψ|(Â ⊗ B̂)|ψi = (hβ| ⊗ hα|)(Â ⊗ B̂)(|αi ⊗ |βi) = hα|Â|αihβ|B̂|βi

= hψ|Â ⊗ IB |ψihψ|IA ⊗ B̂|ψi.

(ii) We use the shortcut notation hÂ1 ⊗ B̂1 i ≡ hψ|Â1 ⊗ B̂1 |ψi etc. Using
statistical independence we have

|hÂ1 ⊗ B̂1 i + hÂ1 ⊗ B̂2 i + hÂ2 ⊗ B̂1 i − hÂ2 ⊗ B̂2 i| =

|hÂ1 ⊗ IB i(hIˆA ⊗ B̂1 i + hIˆA ⊗ B̂2 i) + hÂ2 ⊗ IB i(hIA ⊗ B̂1 i − hIA ⊗ B̂2 i)|.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 298

298 Problems and Solutions

Using the fact that |hÂ1 ⊗ IB i| ≤ 1 and |hÂ2 ⊗ IB i| ≤ 1 we have

|hÂ1 ⊗ B̂1 i + hÂ1 ⊗ B̂2 i + hÂ2 ⊗ B̂1 i − hÂ2 ⊗ B̂2 i|

≤ |hIA ⊗ B̂1 i + hIA ⊗ B̂2 i| + |hIA ⊗ B̂1 i − hIA ⊗ B̂2 i|
≤ max((hIA ⊗ B̂1 i + hIA ⊗ B̂2 i) + (hIA ⊗ B̂1 i − (hIA ⊗ B̂2 i),
(hIA ⊗ B̂1 i + hIA ⊗ B̂2 i) − (hIA ⊗ B̂1 i − hIA ⊗ B̂2 i),
−(hIˆA ⊗ B̂1 i + hIA ⊗ B̂1 i) + (hIA ⊗ B̂1 i − hIA ⊗ B̂2 i),
−(hIA ⊗ B̂1 i + hIA ⊗ B̂2 i) − (hIA ⊗ B̂1 i − hIA ⊗ B̂1 i))
= max(2hIA ⊗ B̂1 i, 2hIA ⊗ B̂2 i, −2hIA ⊗ B̂2 i, −2hIA ⊗ B̂1 i) ≤ 2

where we also used |hIA ⊗ B̂1 i| ≤ 1, |hIA ⊗ B̂2 i| ≤ 1.

Problem 2. Let HA = HB = C2 . Let { |0i , |1i } be the standard basis

in C2 . Consider the entangled state in H = C4 (EPR state)

1
|ψi = √ (|0i ⊗ |1i − |1i ⊗ |0i)
2

which is one of the Bell states. Show that this state and the operators

Â1 := σ1 , Â2 := σ2

1 1
B̂1 := √ (σ1 + σ2 ), B̂2 := √ (σ1 − σ2 )
2 2
violate the Bell inequality.

Solution 2. We have

Â1 |0i = |1i, Â1 |1i = |0i

Â2 |0i = i|1i, Â2 |1i = −i|0i

1 1
B̂1 |0i = √ (|1i + i|1i), B̂1 |1i = √ (|0i − i|0i)
2 2
1 1
B̂2 |0i = √ (|1i − i|1i), B̂2 |1i = √ (|0i − i|0i).
2 2
Using h0|0i = h1|1i = 1 and h0|1i = h1|0i = 0, we find
√
|hψ|Â1 ⊗ B̂1 |ψi + hψ|Â1 ⊗ B̂2 |ψi + hψ|Â2 ⊗ B̂1 |ψi − hψ|Â2 ⊗ B̂2 |ψi| = 2 2.
√
Thus the Bell inequality is violated since 2 2 > 2.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 299

Bell Inequality 299

Problem 3. Let σ1 , σ2 and σ3 be the Pauli spin matrices. Consider the

Bell operator defined by
3
1 X
B := (aj (ck + dk ) + bj (ck − dk ))σj ⊗ σk
2
j,k=1

where a, b, c, d are real unit vectors in R3 .

(i) Calculate the matrix B 2 .
(ii) Consider the Bell state

1
 
+ 1 0
|Φ i = √   .
2 0
1

Calculate the density operator ρ = |Φ+ ihΦ+ | and then tr(ρB). Discuss.
(iii) Let
1
 
0
|e1 i =   .
0
0
Calculate the density operator ρ = |e1 ihe1 | and then tr(ρB). Compare to
(ii).

Solution 3. (i) Since

(σj ⊗ σk )(σm ⊗ σn ) ≡ (σj σm ) ⊗ (σk σn )

we have
3 3
1 X X
B2 = (aj (ck + dk ) + bj (ck − dk ))(am (cn + dn )
4 m,n=1
j,k=1
+bm (cn − dn ))(σj σm ) ⊗ (σk σn ).

Since σj2 = I2 ,

σ1 σ2 = −σ2 σ1 = iσ3 , σ2 σ3 = −σ3 σ2 = iσ1 , σ3 σ1 = −σ1 σ3 = iσ2

and B 2 is hermitian (since B is hermitian) we obtain

B 2 = I2 ⊗ I2 + (a2 b3 − a3 b2 )(c2 d3 − c3 d2 )σ1 ⊗ σ1

+(a1 b3 − b1 a3 )(c1 d3 − d1 c3 )σ2 ⊗ σ2
+(a1 b2 − a2 b1 )(c1 d2 − c2 d1 )σ3 ⊗ σ3
+(a2 b3 − b2 a3 )(c3 d1 − c1 d3 )σ1 ⊗ σ2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 300

300 Problems and Solutions

+(a1 b3 − b1 a3 )(c3 d2 − c2 d3 )σ2 ⊗ σ1

+(a2 b3 − a3 b2 )(c1 d2 − c2 d1 )σ1 ⊗ σ3
+(a1 b2 − a2 b1 )(c2 d3 − c3 d2 )σ3 ⊗ σ1
+(a1 b3 − a3 b1 )(c2 d1 − c1 d2 )σ2 ⊗ σ3
+(a1 b2 − a2 b1 )(c3 d1 − c1 d3 )σ3 ⊗ σ2 .

(ii) We obtain
1 0 0 1
 
1 0 0 0 0
ρ = |Φ+ ihΦ+ | =  .

2 0 0 0 0
1 0 0 1
Thus ρ can be written as a sum of Kronecker products
       
1 1 0 1 0 0 0 0 0
ρ= ⊗ + ⊗
2 0 0 0 0 0 1 0 1
       
0 1 0 1 0 0 0 0
+ ⊗ + ⊗ .
0 0 0 0 1 0 1 0
Since      
1 0 0 0
tr σj = 0, tr σj =0
0 0 0 1
for j = 1, 2 and
     
1 0 0 0
tr σ3 = 1, tr σ3 = −1,
0 0 0 1
     
0 1 0 0
tr σ1 = 1, tr σ1 = 1,
0 0 1 0
     
0 1 0 0
tr σ2 = i, tr σ2 = −i
0 0 1 0
     
0 1 0 0
tr σ3 = 0, tr σ3 = 0
0 0 1 0
we find
1
tr(ρB) = (a1 (c1 + d1 ) + b1 (c1 − d1 ) − a2 (c2 + d2 ) − b2 (c2 − d2 )
2
+a3 (c3 + d3 ) + b3 (c3 − d3 )) .

(iii) We have
1 0 0 0
 
   
0 0 0 0 1 0 1 0
ρ = |e1 ihe1 | =  = ⊗ .
0 0 0 0 0 0 0 0

0 0 0 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 301

Bell Inequality 301

Thus
1
tr(ρB) = (a3 (c3 + d3 ) + b3 (c3 − d3 )).
2

Problem 4. Consider the Pauli spin matrices σ = (σ1 , σ2 , σ3 ). Let q, r,

s, t be unit vectors in R3 . We define

Q := q · σ, R := r · σ, S := s · σ, T := t · σ.

Calculate the matrix (Q ⊗ S + R ⊗ S + R ⊗ T − Q ⊗ T )2 and express the

result using commutators.

Solution 4. Using that for j, k = 1, 2, 3 we have

3
X
σj σk = δjk I2 + i jk` σ`
`=1

with 123 = 231 = 312 = 1, 321 = 213 = 132 = −1 and 0 otherwise we

obtain

(Q ⊗ S + R ⊗ S + R ⊗ T − Q ⊗ T )2 ≡ 4I2 ⊗ I2 + [Q, R] ⊗ [S, T ].

Problem 5. Let X1 and X2 be m × m hermitian matrices with

X12 = X22 = Im .

Let Y1 and Y2 be n × n hermitian matrices with

Y12 = Y22 = In .

(i) What can be said about the eigenvalues of X1 , X2 , Y1 and Y2 ?

(ii) Consider the so-called Bell operator

B := X1 ⊗ (Y1 + Y2 ) + X2 ⊗ (Y1 − Y2 ).

Calculate B 2 . Express the result using commutators.

(iii) What can be said about the eigenvalues of B 2 and B?

Solution 5. (i) From the eigenvalue equation X1 x1 = λx1 we obtain

X1 (X1 x1 ) = X1 (λx1 ) = λX1 x1 = λ2 x1 .

Consequently Im x1 = λ2 x1 and therefore 1 = λ2 . Thus the eigenvalues can

only be ±1. Analogously, we find that the eigenvalues for the matrices X2 ,
Y1 and Y2 can only be ±1.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 302

302 Problems and Solutions

(ii) Since X12 = X22 = Im and Y12 = Y22 = In we have

B 2 = Im ⊗ (Y1 + Y2 )(Y1 + Y2 ) + (X1 X2 ) ⊗ (Y1 + Y2 )(Y1 − Y2 )

+(X2 X1 ) ⊗ (Y1 − Y2 )(Y1 + Y2 ) + Im ⊗ (Y1 − Y2 )(Y1 − Y2 )
= Im ⊗ (2In + Y1 Y2 + Y2 Y1 ) + X1 X2 ⊗ (Y2 Y1 − Y1 Y2 )
+X2 X1 ⊗ (Y1 Y2 − Y2 Y1 ) + Im ⊗ (2In − Y1 Y2 − Y2 Y1 )
= 4Im ⊗ In − (X1 X2 ) ⊗ [Y1 , Y2 ] + (X2 X1 ) ⊗ [Y1 , Y2 ]
= 4Im ⊗ In − [X1 , X2 ] ⊗ [Y1 , Y2 ]
= 4Im ⊗ In + (i[X1 , X2 ]) ⊗ (i[Y1 , Y2 ]).

(iii) Note that the commutator [A, B] of two hermitian d × d matrices A,

B is in general not hermitian. However i[A, B] is hermitian. Thus we can
find an m × m unitary matrix U and an n × n unitary matrix V such that

U (i[X1 , X2 ])U ∗ = diag(α1 , α2 , . . . , αm )

V (i[Y1 , Y2 ])V ∗ = diag(β1 , β2 , . . . , βn )

with αj , βj ∈ R. It follows that

(U ⊗ V )B 2 (U ∗ ⊗ V ∗ ) = 4Im ⊗ In + (U (i[X1 , X2 ])U ∗ ) ⊗ (V (i[Y1 , Y2 ])V ∗ )

= 4Im ⊗ In + diag(α1 , . . . , αm ) ⊗ diag(β1 , . . . , βn ).

The real eigenvalues αj of the hermitian matrix i[X1 , X2 ] are restricted by

−2 ≤ αj ≤ 2 for j = 1, 2, . . . , m. Analogously, the real eigenvalues βk of the
hermitian matrix i[Y1 , Y2 ] are restricted by −2 ≤ βk ≤ 2 for k = 1, 2, . . . , n.
Thus we have −4 ≤ αj βk ≤ 4. The eigenvalues of B 2 are therefore given
by
4 + αj βk , j = 1, 2, . . . , m, k = 1, 2, . . . , n.
It follows that the eigenvalues of B are
p
|λjk | = 4 + αj βk

with j = 1, 2, . . . , m and k = 1, 2, . . . , n.

Problem 6. Let n, m be unit vectors in R3 . Consider the spin singlet

state (entangled state)
       
1 1 0 0 1
|ψi = √ ⊗ − ⊗ .
2 0 1 1 0

(i) Show that the quantum mechanical expectation values E(n, m)

E(n, m) = hψ|(σ · n) ⊗ (σ · m)|ψi

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 303

Bell Inequality 303

is given by

E(n, m) = −m · n = −knk · kmk cos(φ) = − cos(φn,m ) (1)

where φn,m are the angles between the two quantization directions n and
m. We write cos(φn,m ) to indicate that φm,n is the angle between m and
n.
(ii) The CHSH inequality is given by

|E(n, m) − E(n, m0 )| + |E(n0 , m0 ) + E(n0 , m)| ≤ 2. (2)

Insert (1) into (2) and then find the angles, where the inequality is maxi-
mally violated.

Solution 6. (i) Note that σ · n = n1 σ1 + n2 σ2 + n3 σ3 . Straightforward

calculation yields the result.
(ii) We obtain

| cos(φn,m ) − cos(φn,m0 )| + | cos(φn0 ,m0 ) + cos(φn0 ,m )| ≤ 2.

√
The maximal violation is 2 2, achieved by the angles

φn,m0 = 3π/4, φn,m = φn0 ,m0 = φn0 ,m = π/4

√ √
where cos(3π/4) = −1/ 2 and cos(π/4) = 1/ 2. Angles which violate the
inequality (2) are called Bell angles.

Problem 7. Consider four observers: Alice (A), Bob (B), Charlie (C)
and Dora (D) each having one qubit. Every observer is allowed to choose
between two dichotomic observables. Denote the outcome of observer X’s
measurement by Xi (X = A, B, C, D) with i = 1, 2. Under the assumption
of local realism, each outcome can either take the value +1 or −1. The
correlations between the measurement outcomes of all four observers can
be represented by the product Ai Bj Ck Dl , where i, j, k, l = 1, 2. In a lo-
cal realistic theory, the correlation function of the measurement performed
by all four observers is the average of Ai Bj Ck Dl over many runs of the
experiment
Q(Ai Bj Ck Dl ) := hψ|Ai Bj Ck Dl |ψi.
The Mermin-Ardehali-Belinskii-Klyshko inequality is given by

Q(A1 B1 C1 D1 ) − Q(A1 B1 C1 D2 ) − Q(A1 B1 C2 D1 ) − Q(A1 B2 C1 D1 )

−Q(A2 B1 C1 D1 ) − Q(A1 B1 C2 D2 ) − Q(A1 B2 C1 D2 ) − Q(A2 B1 C1 D2 )
−Q(A1 B2 C2 D1 ) − Q(A2 B1 C2 D1 ) − Q(A2 B2 C1 D1 ) + Q(A2 B2 C2 D2 )
+Q(A2 B2 C2 D1 ) + Q(A2 B2 C1 D2 ) + Q(A2 B1 C2 D2 ) + Q(A1 B2 C2 D2 ) ≤ 4.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 304

304 Problems and Solutions

Each observer X measures the spin of each qubit by projecting it either

along nX X
1 or n2 . Every observer can independently choose between two
arbitrary directions. For a four qubit state |ψi, the correlation functions
are thus given by

Q(Ai Bj Ck Dl ) = hψ|(nA B C D
i · σ) ⊗ (nj · σ) ⊗ (nk · σ) ⊗ (nl · σ)|ψi

where · denotes the scalar product, i.e. nX X X X

j · σ := nj1 σ1 + nj2 σ2 + nj3 σ3 .
Let
       
1 0 0 0
nA
1 =  0  , n A
2 =  0  , nB
1 =  1  , nB
2 =  0,
0 1 0 1
       
0 0 −1 1
1 1
nC1 =
 1  , nC 2 =
 0  , nD 1 = √
 0  , nD 2 = √
0.
0 1 2 1 2 1

Show that the Mermin-Ardehali-Belinskii-Klyshko inequality is violated for

the state
1
|ψi = √ (|0000i−|0011i−|0101i+|0110i+|1001i+|1010i+|1100i+|1111i)
2 2
where    
1 0
|0i = , |1i = .
0 1
and |0000i ≡ |0i ⊗ |0i ⊗ |0i ⊗ |0i etc..

Solution 7. For the first term we have

1
Q(A1 B1 C1 D1 ) = √ hψ|σ1 ⊗ σ2 ⊗ σ2 ⊗ (−σ1 + σ3 )|ψi
2
1
= − √ hψ|(σ1 ⊗ σ2 ⊗ σ2 ⊗ σ1 )|ψi
2
1
+ √ hψ|(σ1 ⊗ σ2 ⊗ σ2 ⊗ σ3 )|ψi.
2
Using hψ|(σ1 ⊗ σ2 ⊗ σ2 ⊗ σ1 )|ψi = −1, hψ|(σ1 ⊗ σ2 ⊗ σ2 ⊗ σ3 )|ψi = 0 we
obtain for the first term
1
Q(A1 B1 C1 D1 ) = √ .
2
Analogously√we calculate√the other terms. Summing up the terms we find
the value 4 2. Since 4 2 > 4 the Mermin-Ardehali-Belinskii-Klyshko
inequality is violated by the state |ψi.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 305

Bell Inequality 305

Programming Problem

Problem 1. Let  = σ1 , Â2 = σ2

1 1
B̂1 = √ (σ1 + σ2 ), B̂2 = √ (σ1 − σ2 ).
2 2

(i) Consider the entangled state

T
|ψi = ( 1 1 1 −1 ) .

Is the Bell inequality violated?

(ii) Consider the product state

1
 
   
1 1 1 1 1 1
|φi =   ≡ √ ⊗√ .
2 1 2 1 2 1
1

Is the Bell inequality violated?

Solution 1. We apply the Maxima program

/* Bellinequality.mac */
A1: matrix([0,1],[1,0]); A2: matrix([0,-%i],[%i,0]);
B1: (A1+A2)/sqrt(2); B2: (A1-A2)/sqrt(2);
K11: kronecker_product(A1,B1);
K12: kronecker_product(A1,B2);
K21: kronecker_product(A2,B1);
K22: kronecker_product(A2,B2);
psi: matrix([1],[1],[1],[-1])/2;
psiT: transpose(psi);
S: psiT . K11 . psi + psiT . K12 . psi
+ psiT . K21 . psi - psiT . K22 . psi;
S: ratsimp(S);
Sabs: abs(S);
phi: matrix([1],[1],[1],[1])/2;
phiT: transpose(phi);
S: phiT . K11 . phi + phiT . K12 . phi
+ phiT . K21 . phi - phiT . K22 . phi;
S: ratsimp(S);
Sabs: abs(S);

In both cases the Bell inequality is not violated which is obvious for the
second case since it is a product state.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 306

306 Problems and Solutions

11.3 Supplementary Problems

Problem 1. Let σ1 , σ2 and σ3 be the Pauli spin matrices. Consider the

Bell operator defined by
3
1 X
B := (aj (ck + dk ) + bj (ck − dk ))σj ⊗ σk
2
j,k=1

where a, b, c, d are real unit vectors in R3 .

(i) Consider the orthonormal basis in C4

−1 1 1 1
      
1 1  1  −1  1 1  1 1 
v1 =   , v2 =   , v3 =   , v4 =  .
2 1 2 1 2 −1 2 1
1 1 1 −1

Calculate the density operators ρj = vj vj∗ and then tr(ρj B). Discuss.
(ii) Consider the basis in C4

−1 0 1 1

      
1  1  1  −1  1  0  1  1 
u1 = √   , u2 = √   , u3 = √   , u4 = √  .
3 1 3 1 3 −1 3 0
0 1 1 −1

Calculate the density operators ρj = uj u∗j and tr(ρj B). Compare to (i).

Problem 2. Consider the Hilbert space HA = C2 and the Pauli spin

matrices    
0 1 0 −i
Â1 = σ1 = , Â2 = σ2 =
1 0 i 0
the Hilbert space HB = C3 with the hermitian matrices
   
−1 0 0 0 0 1
B̂1 =  0 1 0  , B̂2 =  0 −1 0 
0 0 −1 1 0 0

with B̂12 = I3 , B̂22 = I3 and the normalized state in C6

1 T
|ψi = √ ( 1 0 0 0 0 −1 ) .
2
Is the Bell inequality violated?
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 307

Chapter 12

Teleportation

12.1 Introduction
Teleportation is the transmission of quantum information using a classical
channel and entanglement. It demonstrates the use of entanglement as a
communication resource. The simplest case is to consider the teleportation
of a single qubit using two bits of classical communication and one entan-
gled pair (EPR-pair). Quantum teleportation is the disembodied transport
of an unknown quantum state from one place to another. The key idea is
that two distant operators, Alice at a sending station and Bob at a receiv-
ing terminal, share an entangled quantum bipartite state and exploit its
nonlocal character as a quantum resource. The resource state can be the
singlet state of a pair of spin- 21 particles (Bell state)

1
|ψi = √ (|0i ⊗ |1i − |1i ⊗ |0i).
2
Particle 1 is given to Alice and particle 2 is given to Bob. Alice intends
to transport an unknown state of a third spin- 12 particle to Bob. She per-
forms a complete projective measurement on the joint system consisting of
particle 1 and 3 and then conveys its outcome to Bob via a classical commu-
nication channel. As a consequence of Alice’s measurement, the total-spin
state of the three-particle system collapses. Owing to the entanglement,
this involves a breakdown of the spin- 21 state of Bob’s particle 2. Never-
theless, Bob makes use of the information transmitted classically by Alice

307
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 308

308 Problems and Solutions

to transform his reduced state into an output that is an accurate replica of

the original unknown input.

12.2 Solved Problems

Problem 1. Consider the following states (a, b ∈ C)
|ψi := a|0i + b|1i, |a|2 + |b|2 = 1 (1)
1
|φi := |ψi ⊗ √ (|00i + |11i). (2)
2
(i) Show that the state |φi can be written as
1 1
|φi = √ (|00i + |11i) ⊗ (a|0i + b|1i) + √ (|00i − |11i) ⊗ (a|0i − b|1i)
2 2 2 2
1 1
+ √ (|01i + |10i) ⊗ (a|1i + b|0i) + √ (|01i − |10i) ⊗ (a|1i − b|0i) .
2 2 2 2
(ii) Describe how measurement of the first two qubits of |φi can be used to
obtain |ψi as the last qubit. Alice has the first qubit of |φi and Alice and
Bob share the second and third qubits of |φi (an EPR-pair).

Solution 1. (i) Inserting (1) into (2) we obtain

1
|φi = √ (a|000i + a|011i + b|100i + b|111i) .
2
On the other hand we have
1 1
√ (|00i + |11i) ⊗ (a|0i + b|1i) + √ (|00i − |11i) ⊗ (a|0i − b|1i)
2 2 2 2
1 1
+ √ (|01i + |10i) ⊗ (a|1i + b|0i) + √ (|01i − |10i) ⊗ (a|1i − b|0i)
2 2 2 2
1
= √ (a|000i + a|110i + b|001i + b|111i)
2 2
1
+ √ (a|000i − a|110i − b|001i + b|111i)
2 2
1
+ √ (a|011i + a|101i + b|010i + b|100i)
2 2
1
+ √ (a|011i − a|101i − b|010i + b|100i)
2 2
1
= √ (a|000i + a|011i + b|100i + b|111i)
2
= |φi.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 309

Teleportation 309

(ii) We measure in the Bell basis

 
1 1 1 1
√ (|00i + |11i), √ (|00i − |11i), √ (|01i + |10i), √ (|01i − |10i) .
2 2 2 2

From the state |φi we can see that the first two qubits are in each of the
Bell states with equal probability. Thus if we measure the first two qubits
in the Bell basis we obtain a result corresponding to each of the Bell states
and can perform a transform to obtain |ψi in the last qubit as follows

Bell State Transform

√1 (|00i + |11i) I2
2
√1 (|00i − |11i) |0ih0| − |1ih1|
2
√1 (|01i + |10i) UN OT
2
√1 (|01i − |10i) |0ih1| − |1ih0|
2

After measurement and applying the corresponding transform we obtain

|ψi as the last qubit. So if Alice and Bob initially share the entangled pair

1 1
√ (|00i + |11i) ≡ √ (|0iA ⊗ |0iB + |1iA ⊗ |1iB ).
2 2
Alice can perform a measurement in the Bell basis on her qubit and her part
of the entangled pair and sends the result (two bits) to Bob who applies
the corresponding transform to his part of the entangled pair. The state
|ψi is thus teleported from Alice’s qubit to Bob’s qubit. Note that the Bell
basis is obtained by applying UCN OT (UH ⊗ I2 ) to the computational basis
{|00i, |01i, |10i, |11i}. The transforms are unitary and therefore invertible.
Thus we can also measure the first two qubits in the computational basis
after applying the unitary matrix

(UH ⊗ I2 )UCN OT .

Problem 2. Consider the state in the Hilbert space H = C16

|ψ0 i = |0101i ≡ |0i ⊗ |1i ⊗ |0i ⊗ |1i

where { |0i, |1i } is the standard basis in C2 . Let

1
|ψ1 i = B|ψ0 i = √ (|0101i + |0110i)
2
1
|ψ2 i = U |ψ1 i = √ (|0101i + |1010i)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 310

310 Problems and Solutions

1
|ψ3 i = S|ψ2 i = √ (|0101i − |1010i)
2
1
|ψ4 i = U ∗ |ψ3 i = √ (|0101i − |0110i)
2
|ψ5 i = B ∗ |ψ4 i = −|0110i.

Find the 16 × 16 unitary matrices B, U , S which perform these transfor-

mations.

Solution 2. From the above equations we find the following

1
B|0101i = √ (|0101i + |0110i)
2
1 1
U √ (|0101i + |0110i) = √ (|0101i + |1010i)
2 2
1 1
S √ (|0101i + |1010i) = √ (|0101i − |1010i)
2 2
∗ 1 1
U √ (|0101i − |1010i) = √ (|0101i − |0110i)
2 2
1
B ∗ √ (|0101i − |0110i) = −|0110i.
2

A unitary transform maps an orthonormal basis to an orthonormal basis.

The above equations do not determine B, U and S uniquely. For simplicity
let B, U and S act as the identity on subspaces for which the unitary
transformations are not constrained by the above equations. For B we
have
1 1
B|0101i = |01i ⊗ √ (|01i + |10i), B|0110i = |01i ⊗ √ (|10i − |01i) .
2 2

One solution is
1
B = √ I4 ⊗ (|γih01| + |δih10| + |αih00| + |βih11|)
2

where

|αi = |00i + |11i, |βi = |00i − |11i, |γi = |01i + |10i, |δi = |10i − |01i.

This means that B maps from the computational basis to the Bell basis in
the second two qubits. For U we have
 
1 1
U |01i ⊗ √ (|01i + |10i) = √ (|0101i + |1010i)
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 311

Teleportation 311

and  
1 1
U |01i ⊗ √ (|01i − |10i) = √ (|0101i − |1010i).
2 2
We rewrite these equations in the simpler form

U |0101i = |0101i, U |0110i = |1010i.

A solution for U is then

U = I16 + (|1010i − |0110i)(h0110| − h1010|)

i.e. U is the identity except on the subspace spanned by |0110i and |1010i,
where U swaps |0110i and |1010i. For S we have
1 1
S √ (|0101i + |1010i) = √ (|0101i − |1010i).
2 2
A solution for S is S = I16 − 2|1010ih1010| i.e. S is the identity except for
changing the sign of |1010i.

Problem 3. Let |ψi := a|0i + b|1i be an arbitrary qubit state. Let |φi
be another arbitrary qubit state. Let U be a unitary operator which acts
on two qubits.
(i) Determine the implications of measuring the first two qubits of
1
|θi := |ψi ⊗ √ (I2 ⊗ U ) ((|00i + |11i) ⊗ |φi)
2
with respect to the Bell basis. How can we obtain U (|ψi ⊗ |φi) as the last
two qubits?
(ii) Alice has |ψi and Bob has |φi. Describe how U can be applied to |ψi⊗|φi
using only classical communication and prior shared entanglement. After
the computation, Alice must still have the first qubit of U (|ψi ⊗ |φi) and
Bob must still have the second qubit of U (|ψi ⊗ |φi).

Solution 3. (i) We have

1
|θi = a|0i ⊗ √ (I2 ⊗ U ) ((|00i + |11i) ⊗ |φi)
2
1
+b|1i ⊗ √ (I2 ⊗ U ) ((|00i + |11i) ⊗ |φi)
2
1
= a|0i ⊗ √ (|0i ⊗ U (|0i ⊗ |φi) + |1i ⊗ U (|1i ⊗ |φi))
2
1
+b|1i ⊗ √ (|0i ⊗ U (|0i ⊗ |φi) + |1i ⊗ U (|1i ⊗ |φi))
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 312

312 Problems and Solutions

1
= √ (|00i ⊗ U (a|0i ⊗ |φi) + |01i ⊗ U (a|1i ⊗ |φi))
2
1
+ √ (|10i ⊗ U (b|0i ⊗ |φi) + |11i ⊗ U (b|1i ⊗ |φi)).
2
Expanding |00i, |01i, |10i and |11i in the Bell basis for the first two qubits
yields the state
1
|θi = √ (|00i + |11i) ⊗ U ((a|0i + b|1i) ⊗ |φi)
2 2
1
+ √ (|00i − |11i) ⊗ U ((a|0i − b|1i) ⊗ |φi)
2 2
1
+ √ (|01i + |10i) ⊗ U ((a|1i + b|0i) ⊗ |φi)
2 2
1
+ √ (|01i − |10i) ⊗ U ((a|1i − b|0i) ⊗ |φi) .
2 2
We measure in the Bell basis
 
1 1 1 1
√ (|00i + |11i), √ (|00i − |11i), √ (|01i + |10i), √ (|01i − |10i) .
2 2 2 2
From |θi we can see that the first two qubits are in each of the Bell states
with equal probability. If we make a measurement we obtain a result corre-
sponding to each of the Bell states and can perform a transform to obtain
U (|ψi ⊗ |φi) in the last two qubits as follows
Bell State Transform
√1 (|00i + |11i) I2
2
√1 (|00i
2
− |11i) U ((|0ih0| − |1ih1|) ⊗ I2 ) U ∗
√1 (|01i
2
+ |10i) U (UN OT ⊗ I2 )U ∗
√1 (|01i
2
− |10i) U ((|0ih1| − |1ih0|) ⊗ I2 ) U ∗

Thus after measurement and applying the corresponding transform we ob-

tain U (|ψi ⊗ |φi) as the last two qubits. Thus if Alice and Bob initially
share the entangled state
1
√ (|00i + |11i)
2
Bob applies U to his two qubits. Then Alice can perform a measurement
in the Bell basis on her qubit and her part of the entangled pair and sends
the result (two bits) to Bob who applies the corresponding transform from
the table to his part of the entangled pair. Thus with probability 14 Bob
can begin the computation U (|ψi ⊗ |φi) without knowing the state |ψi and
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 313

Teleportation 313

still obtain the correct result after Alice measures her two qubits. With
probability 34 he still has to apply a transform which is independent of |ψi.
(ii) Alice teleports |ψi to Bob with one entangled pair, Bob performs the
computation U (|ψi⊗|φi) on his two qubits and then teleports the first qubit
back to Alice with a second entangled pair. Thus 4 bits of communication
are used in this scheme (Alice sends two to Bob, and then Bob sends two
to Alice). Alice and Bob can perform UCN OT even though their qubits are
spatially separated if they have prior entanglement.

Problem 4. In quantum teleportation we start with the following state

in the Hilbert space C8

|ψi ⊗ |0i ⊗ |0i ≡ (a|0i + b|1i) ⊗ |0i ⊗ |0i ≡ |ψ00i

where |a|2 + |b|2 = 1. The quantum circuit for teleportation is given by

A r H r

B H r g r

C g g H g H

where A is the input |ψi, B the input |0i and C the input |0i. Study what
happens when we feed the product state |ψ00i into the quantum circuit.
From the circuit we have the following eight 8 × 8 unitary matrices (left to
right)

U1 = I2 ⊗ UH ⊗ I2 , U2 = I2 ⊗ UXOR ,
U3 = UXOR ⊗ I2 , U4 = UH ⊗ I2 ⊗ I2 ,
U5 = I2 ⊗ UXOR , U6 = I2 ⊗ I2 ⊗ UH ,
U7 = I4 ⊕ UN OT ⊕ UN OT , U8 = I2 ⊗ I2 ⊗ UH

where ⊕ denotes the direct sum of matrices, UH denotes the Hadamard

gate, UXOR denotes the XOR-gate and
 
0 1
UN OT := .
1 0

(i) Find U8 U7 U6 U5 U4 U3 U2 U1 |ψ00i.

(ii) Write a program which implements and verifies the teleportation algo-
rithm.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 314

314 Problems and Solutions

Solution 4. (i) Applying the first four unitary matrices to the input state
we obtain

U4 U3 U2 U1 |ψ00i
a b
= (|000i + |100i + |011i + |111i) + (|010i − |110i + |001i − |101i).
2 2
This state can be rewritten as
1 a
U4 U3 U2 U1 |ψ00i = √ (|0i + |1i) ⊗ ( √ (|0i ⊗ |0i + |1i ⊗ |1i))
2 2
1 b
+ √ (|0i − |1i) ⊗ ( √ (|0i ⊗ |1i + |1i ⊗ |0i)).
2 2
Applying all eight unitary matrices to the input state we obtain

U8 U7 U6 U5 U4 U3 U2 U1 |ψ00i
a b
= (|000i + |100i + |010i + |110i) + (|011i + |111i + |001i + |101i).
2 2
This state can be rewritten as
   
1 1
√ (|0i + |1i) ⊗ √ (|0i + |1i) ⊗ |ψi.
2 2
The state |ψi will be transferred to the lower √ output, where both other
outputs will come out in the state (|0i + |1i)/ 2. If the two upper outputs
are measured in the standard basis (|0i versus |1i), two random classical
bits will be obtained in addition to the quantum state |ψi on the lower
output.
(ii) The implementation in SymbolicC++ is as follows. The Matrix class
of SymbolicC++ includes the method kron for the Kronecker product of
two matrices and the method dsum for the direct sum of two matrices.
The overloaded operators * and + are used for matrix multiplication and
addition. The identity matrix is also implemented. Thus the code for the
quantum circuit is as follows.
// teleport.cpp
#include <iostream>
#include "symbolicc++.h"
using namespace std;

Symbolic Hadamard(const Symbolic &v)

{
Symbolic H("",2,2);
Symbolic sqrt12 = sqrt(1/Symbolic(2));
H(0,0) = sqrt12; H(0,1) = sqrt12;
H(1,0) = sqrt12; H(1,1) = -sqrt12;
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 315

Teleportation 315

return (H*v);
}

Symbolic XOR(const Symbolic &v)

{
Symbolic X("",4,4);
X(0,0) = 1; X(0,1) = 0; X(0,2) = 0; X(0,3) = 0;
X(1,0) = 0; X(1,1) = 1; X(1,2) = 0; X(1,3) = 0;
X(2,0) = 0; X(2,1) = 0; X(2,2) = 0; X(2,3) = 1;
X(3,0) = 0; X(3,1) = 0; X(3,2) = 1; X(3,3) = 0;
return (X*v);
}

Symbolic Bell(const Symbolic &v)

{
Symbolic I("",2,2), H("",2,2), X("",4,4);
Symbolic sqrt12 = sqrt(1/Symbolic(2));
I = I.identity();
H(0,0) = sqrt12; H(0,1) = sqrt12;
H(1,0) = sqrt12; H(1,1) = -sqrt12;
Symbolic UH = kron(H,I);
X(0,0) = 1; X(0,1) = 0; X(0,2) = 0; X(0,3) = 0;
X(1,0) = 0; X(1,1) = 1; X(1,2) = 0; X(1,3) = 0;
X(2,0) = 0; X(2,1) = 0; X(2,2) = 0; X(2,3) = 1;
X(3,0) = 0; X(3,1) = 0; X(3,2) = 1; X(3,3) = 0;
return (X*(UH*v));
}

Symbolic Swap(const Symbolic &v)

{
Symbolic S("",4,4);
S(0,0) = 1; S(0,1) = 0; S(0,2) = 0; S(0,3) = 0;
S(1,0) = 0; S(1,1) = 0; S(1,2) = 0; S(1,3) = 1;
S(2,0) = 0; S(2,1) = 0; S(2,2) = 1; S(2,3) = 0;
S(3,0) = 0; S(3,1) = 1; S(3,2) = 0; S(3,3) = 0;
return XOR(S*XOR(v));
}

Symbolic Teleport(const Symbolic &v)

{
Symbolic result;
Symbolic NOT("",2,2),H("",2,2),I("",2,2),X("",4,4);
Symbolic sqrt12 = sqrt(1/Symbolic(2));
NOT(0,0) = 0; NOT(0,1) = 1; NOT(1,0) = 1; NOT(1,1) = 0;
H(0,0) = sqrt12; H(0,1) = sqrt12;
H(1,0) = sqrt12; H(1,1) = -sqrt12;
I = I.identity();
X(0,0) = 1; X(0,1) = 0; X(0,2) = 0; X(0,3) = 0;
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 316

316 Problems and Solutions

X(1,0) = 0; X(1,1) = 1; X(1,2) = 0; X(1,3) = 0;

X(2,0) = 0; X(2,1) = 0; X(2,2) = 0; X(2,3) = 1;
X(3,0) = 0; X(3,1) = 0; X(3,2) = 1; X(3,3) = 0;
Symbolic U1 = kron(I,kron(H,I)); Symbolic U2 = kron(I,X);
Symbolic U3 = kron(X,I); Symbolic U4 = kron(H,kron(I,I));
Symbolic U5 = kron(I,X); Symbolic U6 = kron(I,kron(I,H));
Symbolic U7 = dsum(I,dsum(I,dsum(NOT,NOT)));
Symbolic U8 = kron(I,kron(I,H));
result = U8*(U7*(U6*(U5*(U4*(U3*(U2*(U1*v)))))));
return result;
}

// The outcome after measuring value for qubit.

// Since the probabilities may be symbolic this function
// cannot simulate a measurement where random outcomes
// have the correct distribution
Symbolic Measure(const Symbolic &v,unsigned int qubit,
unsigned int value)
{
int i,len,skip = 1-value;
Symbolic result(v);
Symbolic D;
len = v.rows()/int(pow(2.0,qubit+1.0));
for(i=0;i<v.rows();i++)
{
if(!(i%len)) skip = 1-skip;
if(skip) result(i) = 0; else D += result(i)*result(i);
}
return result/sqrt(D);
}

// for output clarity

ostream &print(ostream &o,const Symbolic &v)
{
char *b2[2]={"|0>","|1>"};
char *b4[4]={"|00>","|01>","|10>","|11>"};
char *b8[8]={"|000>","|001>","|010>","|011>",
"|100>","|101>","|110>","|111>"};
char **b;
if(v.rows()==2) b=b2;
if(v.rows()==4) b=b4;
if(v.rows()==8) b=b8;
for(int i=0;i<v.rows();i++)
if(v(i)!=0) o << "+(" << v(i) << ")" << b[i];
return o;
}

int main(void)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 317

Teleportation 317

{
Symbolic zero("",2),one("",2);
Symbolic zz("",4),zo("",4),oz("",4),oo("",4),qreg;
Symbolic tp00,tp01,tp10,tp11,psiGHZ;
Symbolic a("a"), b("b");
Symbolic sqrt12 = sqrt(1/Symbolic(2));
zero(0) = 1; zero(1) = 0; one(0) = 0; one(1) = 1;
zz = kron(zero,zero); zo = kron(zero,one);
oz = kron(one,zero); oo = kron(one,one);
cout << "UH|0> = "; print(cout,Hadamard(zero))<< endl;
cout << "UH|1> = "; print(cout,Hadamard(one)) << endl;
cout << endl;
cout << "UXOR|00> = "; print(cout,XOR(zz)) << endl;
cout << "UXOR|01> = "; print(cout,XOR(zo)) << endl;
cout << "UXOR|10> = "; print(cout,XOR(oz)) << endl;
cout << "UXOR|11> = "; print(cout,XOR(oo)) << endl;
cout << endl;
cout << "UBELL|00> = "; print(cout,Bell(zz)) << endl;
cout << "UBELL|01> = "; print(cout,Bell(zo)) << endl;
cout << "UBELL|10> = "; print(cout,Bell(oz)) << endl;
cout << "UBELL|11> = "; print(cout,Bell(oo)) << endl;
cout << endl;
cout << "USWAP|00> = "; print(cout,Swap(zz)) << endl;
cout << "USWAP|01> = "; print(cout,Swap(zo)) << endl;
cout << "USWAP|10> = "; print(cout,Swap(oz)) << endl;
cout << "USWAP|11> = "; print(cout,Swap(oo)) << endl;
cout << endl;
qreg=kron(a*zero+b*one,kron(zero,zero));
cout << "UTELEPORT("; print(cout,qreg) << ") = ";
print(cout,qreg=Teleport(qreg)) << endl;
cout << "Results after measurement of first 2 qubits:" << endl;
tp00 = Measure(Measure(qreg,0,0),1,0);
tp01 = Measure(Measure(qreg,0,0),1,1);
tp10 = Measure(Measure(qreg,0,1),1,0);
tp11 = Measure(Measure(qreg,0,1),1,1);
Equations simplify = (a*a==1-b*b,1/sqrt(1/Symbolic(4))==2);
tp00 = tp00.subst_all(simplify);
tp01 = tp01.subst_all(simplify);
tp10 = tp10.subst_all(simplify);
tp11 = tp11.subst_all(simplify);
cout << " |00> : " ; print(cout,tp00) << endl;
cout << " |01> : " ; print(cout,tp01) << endl;
cout << " |10> : " ; print(cout,tp10) << endl;
cout << " |11> : " ; print(cout,tp11) << endl;
cout << endl;
psiGHZ=kron(zz,zero)*sqrt12+kron(oo,one)*sqrt12;
cout << "Greenberger-Horne-Zeilinger state : ";
print(cout,psiGHZ) << endl;
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 318

318 Problems and Solutions

cout << "Measuring qubit 0 as 1 yields : ";

print(cout,Measure(psiGHZ,0,1)) << endl;
cout << "Measuring qubit 1 as 1 yields : ";
print(cout,Measure(psiGHZ,1,1)) << endl;
cout << "Measuring qubit 2 as 0 yields : ";
print(cout,Measure(psiGHZ,2,0)) << endl;
return 0;
}
The program generates the following output
UTELEPORT(+(a)|000>+(b)|100>) =
+(1/2*a)|000>+(1/2*b)|001>+(1/2*a)|010>
+(1/2*b)|011>+(1/2*a)|100>+(1/2*b)|101>
+(1/2*a)|110>+(1/2*b)|111>
Results after measurement of first 2 qubits:
|00> : +(a)|000>+(b)|001>
|01> : +(a)|010>+(b)|011>
|10> : +(a)|100>+(b)|101>
|11> : +(a)|110>+(b)|111>

12.3 Supplementary Problems

Problem 1. Let v1 , v2 , v3 be normalized states in C2 . Find the unitary

8 × 8 matrices U1 , U2 , U3 such
U1 (v1 ⊗ v2 ⊗ v3 ) = v2 ⊗ v3 ⊗ v1
U2 (v2 ⊗ v3 ⊗ v1 ) = v3 ⊗ v1 ⊗ v2
U3 (v3 ⊗ v1 ⊗ v2 ) = v1 ⊗ v2 ⊗ v3 .
Discuss

Problem 2. Let |0i, |1i be the standard basis in C2 . Construct a unitary

8 × 8 matrix U applying the generalized Gram-Schmidt technique such that
  
cos(θ)
U ⊗ |0i ⊗ |0i =
sin(θ)
 
1 cos(θ) 1
√ (|0i + |1i) ⊗ ⊗ √ (|0i + |1i).
2 sin(θ) 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 319

Chapter 13

Cloning

13.1 Introduction
Cloning is the duplication of information. Cloning is necessarily a physical
process. Obviously we have to find unitary transformations for the cloning
process. The no cloning theorem is a result of quantum mechanics which
forbids the creation of identical copies of an arbitrary unknown quantum
state. However, approximate copies still have many uses in quantum com-
puting. Owing to the superposition principle of quantum mechanics it is not
possible in general to clone an arbitrary quantum state. As a consequence
any successful attempt to clone a state will destroy the original state in the
process. Quantum cloning machines are devices for approximately cloning
arbitrary quantum states. Suppose we want to clone the state |ψi ∈ H to
obtain
|ψi ⊗ |ψi.
The approximate cloning operation yields some mixed state ρ in the product
Hilbert space H ⊗ H. The fidelity of the cloning process for |ψi is given by

(hψ| ⊗ hψ|)ρ(|ψi ⊗ |ψi).

Optimal cloning attempts to maximize the average fidelity (closeness to

|ψi ⊗ |ψi). In the case of symmetric cloning we assume

hψ|trH (ρ)|ψi

319
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 320

320 Problems and Solutions

is independent of the system on which the partial trace is performed. Thus

for symmetric cloning we need only use hψ|trH (ρ)|ψi as the measure of
fidelity. We provide exercises describing what types of information can be
cloned accurately and techniques for cloning certain types of information.

13.2 Solved Problems

Problem 1. Let a ∈ {0, 1} and |0i, |1i be the standard basis. Find a
unitary 4 × 4 matrix that maps

U : |ai ⊗ |ai 7→ |ai ⊗ |ai

where a denotes the NOT operation applied to a.

Solution 1. Obviously the permutation matrix

0 1 0 0
 
   
1 0 0 0 0 1 0 1
U = ≡ ⊕
0 0 0 1 1 0 1 0
0 0 1 0
provides such a map, where ⊕ denotes the direct sum.

Problem 2. The CNOT gate maps (a, b ∈ {0, 1})

|ai ⊗ |bi → |ai ⊗ |a ⊕ bi

where ⊕ is the XOR operation. Show that the CNOT gate can be used to
clone a bit.

Solution 2. Setting b = 0 we obtain from the CNOT gate

|ai ⊗ |0i → |ai ⊗ |ai

since a ⊕ 0 = a for all a. Thus we have cloned a bit.

Problem 3. Let
 
x1
x= , x1 x∗1 + x2 x∗2 = 1
x2

be an arbitrary normalized vector in C2 . Can we construct a 4 × 4 unitary

matrix U such that
       
x1 1 x1 x1
U ⊗ = ⊗ ? (1)
x2 0 x2 x2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 321

Cloning 321

Prove or disprove this equation.

Solution 3. Such a matrix does not exist. This can be seen as follows.
From the right-hand side of (1) we have
           
x1 x1 x1 0 x1 0
⊗ = + ⊗ + =
x2 x2 0 x2 0 x2
               
x1 x1 x1 0 0 x1 0 0
⊗ + ⊗ + ⊗ + ⊗ .
0 0 0 x2 x2 0 x2 x2
On the other hand, from the left-hand side of (1) we find
           
x1 1 x1 1 0 1
U ⊗ =U ⊗ + ⊗
x2 0 0 0 x2 0
       
x1 x1 0 0
= ⊗ + ⊗
0 0 x2 x2

where we used the linearity of the unitary matrix U . Comparing these two
equations we find a contradiction. This is the no cloning theorem.

However equation (1) does hold when

           
x1 0 0 x1 0 0
⊗ + ⊗ = ⊗ .
0 x2 x2 0 0 0

Therefore x1 x2 = 0. Thus at least one of x1 and x2 must be zero. It is still

possible to clone elements of a known orthonormal basis.

Problem 4. Let |ψi, |si, |φi be normalized states in a Hilbert space H.

Let U be a unitary operator, i.e. U −1 = U ∗ in the product Hilbert space
H ⊗ H such that

U (|ψi ⊗ |si) = |ψi ⊗ |ψi

U (|φi ⊗ |si) = |φi ⊗ |φi.

Show that hφ|ψi = hφ|ψi2 . Find solutions to this equation.

Solution 4. Taking the scalar product of these two equations with U ∗ =

U −1 and hs|si = 1 we obtain

(hψ| ⊗ hs|)U ∗ U (|φi ⊗ |si) = (hψ| ⊗ hψ|)(|φi ⊗ |φi)

(hψ| ⊗ hs|)(|φi ⊗ |si) = hψ|φihψ|φi
hψ|φi = hψ|φi2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 322

322 Problems and Solutions

The equation can be satisfied if hψ|φi = 0 (|ψi and |φi are orthonormal to
each other) or hψ|φi = 1, i.e. |ψi = |φi.

Problem 5. Let { |0i , |1i } be a basis in C2 . Let |ψi be an arbitrary

qubit. Is there a unitary transformation such that
|ψi ⊗ |ψi → |ψi ⊗ |0i ?

Solution 5. Such a unitary transformation does not exist. For an ar-

bitrary qubit |ψi the product states |ψi ⊗ |ψi span a three-dimensional
subspace of the four-dimensional Hilbert space C4 of two qubits. However,
the product states |ψi ⊗ |0i span only a two-dimensional subspace, as |0i
is a fixed state. Thus the unitary transform would take a system with von
Neumann entropy log2 (3) to one with von Neumann entropy log2 (2). Since
the system is closed (we have a unitary transformation), this decrease of en-
tropy is therefore a violation of the second law of thermodynamics. Thus the
second law of thermodynamics implies that such a unitary transformation
does not exist.

Problem 6. Consider the approximate cloning of a qubit |ψi ∈ C2 by

simply measuring the qubit with respect to a randomly chosen orthonormal
basis { |0i, |1i }, and then using the state |0i ⊗ |0i or |1i ⊗ |1i corresponding
to the measurement outcome. Determine the average fidelity of this cloning
process.

Solution 6. The state |ψi is represented by a density operator

1
|ψihψ| = (I2 + n · σ)
2
where n is a unit vector in R3 and n·σ := n1 σ1 +n2 σ2 +n3 σ3 . The randomly
chosen orthonormal basis can be represented by the density operators
1
|0ih0| = (I2 + m · σ)
2
1
|1ih1| = I2 − |0ih0| = (I2 − m · σ)
2
where m is a unit vector in R3 . Measuring |ψi in the basis { |0i, |1i } yields
|0i with probability
1
|h0|ψi|2 = (1 + n · m)
2
and |1i with probability
1
|h1|ψi|2 = (1 − n · m)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 323

Cloning 323

where |h0|ψi|2 = h0|ψihψ|0i. Thus we construct a two qubit system de-

scribed by the density operator

1 1
(1 + n · m)|0ih0| ⊗ |0ih0| + (1 − n · m)|1ih1| ⊗ |1ih1|.
2 2
If we trace out either qubit we are left with the same density operator, i.e.
this is symmetric cloning. The fidelity for |ψi (symmetric cloning) is given
by
 2  2
1 1 1
(1 + n · m) + (1 − n · m) = (1 + (n · m)2 ).
2 2 2
To determine the average fidelity we integrate over all m ∈ R3 with kmk =
1. Thus we use spherical coordinates

m = (sin(θ) cos(φ), sin(θ) sin(φ), cos(θ)), φ ∈ [0, 2π], θ ∈ [0, π].

Then
n · m = n1 sin(θ) cos(φ) + n2 sin(θ) sin(φ) + n3 cos(θ).
Consequently the average fidelity is given by
Z
1 1
(1 + (n · m)2 ) dm
4π kmk=1 2
Z π Z 2π
1 1
= (1 + n · m)2 ) sin(θ)dφ dθ
4π 0 0 2
1 1 π 2 3
Z
= + (n1 sin (θ) + n22 sin3 (θ) + 2n23 cos2 (θ) sin(θ))dθ
2 8 0
1 1 π
Z
= + (2 cos2 (θ) sin(θ) + (n21 + n22 )(sin(θ) − 3 cos2 (θ) sin(θ))) dθ
2 8 0
2
=
3

where we used n23 = 1 − n21 − n22 . The average fidelity is independent of

|ψi, i.e. the cloning process in universal. However the cloning process is
not optimal.

Programming Problem

Problem 1. Show that there is a unitary 4 × 4 matrix such that

        
1 1 1 1 1 1 1 1
U √ ⊗√ =√ ⊗√ .
2 1 2 −1 2 1 2 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 324

324 Problems and Solutions

Solution 1. The unitary matrix

   
1 0 0 −1
U= ⊗
0 1 1 0
will do the job.
/* cloning.mac */
v1: matrix([1],[1])/sqrt(2); v2: matrix([1],[-1])/sqrt(2);
I2: matrix([1,0],[0,1]);
V: matrix([0,-1],[1,0]);
U: kronecker_product(I2,V);
v1v2k: kronecker_product(v1,v2);
v1v2kU: U . v1v2k;
v1v1k: kronecker_product(v1,v1);
if(v1v1k=v1v2kU) then print("U does the cloning")
else print("U does not do the cloning");

13.3 Supplementary Problems

Problem 1. Let a, b ∈ {0, 1}.

(i) Can one find a unitary 4 × 4 matrix such that

U (|ai ⊗ |bi) = |ai ⊗ |a · bi

for all a, b ∈ {0, 1}? Here · denotes the AND operation.

(ii) Can one find a unitary 4 × 4 matrix such that

U (|ai ⊗ |bi) = |ai ⊗ |a + bi

for all a, b ∈ {0, 1}? Here + denotes the OR operation.

Problem 2. Let
 
x1
x =  x2  , x1 x∗1 + x2 x∗2 + x3 x∗3 = 1
x3

be an arbitrary normalized vector in C2 . Can we construct a 9 × 9 unitary

matrix U such that
       
x1 1 x1 x1
U  x2  ⊗  0  =  x2  ⊗  x2  ? (1)
x3 0 x3 x3
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 325

Chapter 14

Quantum Algorithms

14.1 Introduction
An algorithm is a precise description of how to realize a given objective,
for example solving a computational problem. We distinguish between
classical and quantum algorithms where quantum physical resources are
used. Quantum algorithms run on quantum computers and thus utilizing
unitary transformations such as the quantum Fourier transform and super-
position. Some quantum algorithms also utilize entanglement. Quantum
algorithms include: quantum counting, quantum phase estimation algo-
rithm, Deutsch’s algorithm, Deutsch-Josza algorithm, Simon’s algorithm,
Shor’s algorithm, Gover’s algorithm and hidden subgroup problem. Quan-
tum counting algorithm is a quantum algorithm for counting the number
of solutions for a given search problem. The quantum phase estimation
algorithm is used to find the eigenphase of a normalized eigenvector of a
unitary gate (unitary matrix) given a quantum state proportional to the
normalized eigenvector and access to the unitary gate. Deutsch’s problem
leads to the simplest quantum algorithm. Consider the Boolean functions f
that map { 0, 1 } to { 0, 1 }. There are exactly four such functions: two con-
stant functions (f (0) = f (1) = 0 and f (0) = f (1) = 1) and two balanced
functions (f (0) = 0, f (1) = 1 and f (0) = 1, f (1) = 0). In Deutsch’s prob-
lem one is allowed to evaluate the function only once and we are required
to deduce from the result whether f is constant or balanced. Thus we are
asked for the global property of f . A generalization is the Deutsch-Josza

325
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 326

326 Problems and Solutions

problem. One considers the boolean functions

f : { 0, 1 }n → { 0, 1 }

in the following way. Assume that, for one of these functions, it is promised
that it either constant or balanced (i.e. has an equal number of 0’s outputs
as 1’s) and consider the goal of determining which of the two properties
the function actually has. Given a boolean function f : {0, 1}n → {0, 1}.
Assume that f is known to be invariant under some n-bit XOR mask b.
Simon’s problem is to determine b. Shor’s algorithm solves the discrete log-
arithm problem and the integer factorization problem. Grover’s algorithm
searches an unsorted database with N entries for a marked entry. The
algorithms of Deutsch, Simon, Shor and others can be formulated group
theoretically as a hidden subgroup problem. Let f be a function from a
finitely generated group G to a finite set such that f is constant on the
cosets of a subgroup K and distinct on each on each coset. The cosets of
K are the sets
g · K := { g · k : k ∈ K }, g ∈ G.

The cosets partition the group G, i.e. the union of all cosets is the set of
the group G and every two cosets are equal or their intersection is empty.
Thus one writes

K = { k ∈ G : f (k · g) = f (g), ∀ g ∈ G }.

The problem is, for a given f and G determine the hidden subgroup K.

The quantum Fourier transform plays an important role in a number of

quantum algorithms. The quantum Fourier transform on the additive group
of integers modulo 2m is the mapping

m
2X −1
F2m
|ai → exp(2πiay/2m )|yi
y=0

where a ∈ { 0, 1, . . . , 2m − 1 }. The state on the right-hand side is unentan-

gled. The quantum Fourier transform is used in Shor’s algorithm. Given
positive integers a and N which are relatively prime and such that a < N .
The goal is to find the minimum positive integer r such that ar mod N = 1.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 327

Quantum Algorithms 327

14.2 Solved Problems

Problem 1. In classical communication complexity Alice is provided with
a binary string
x = x0 x1 · · · xn−1
of length n and Bob is provided with a binary string

y = y0 y1 · · · yn−1

of length n. Alice has to determine a boolean function

f : { 0, 1 }n × { 0, 1 }n → { 0, 1 }

with the least communication between herself and Bob.

(i) Consider the parity function

f (x, y) = x0 ⊕ x1 ⊕ · · · ⊕ xn−1 ⊕ y0 ⊕ y1 ⊕ · · · ⊕ yn−1

where ⊕ is the XOR operation, i.e.

0 ⊕ 0 = 0, 0 ⊕ 1 = 1, 1 ⊕ 0 = 1, 1 ⊕ 1 = 0.

How many bits has Bob to send to Alice so that she can determine f ?
(ii) Consider the inner product modulo-2 function

f (x, y) = (x0 · y0 ) ⊕ (x1 · y1 ) ⊕ · · · ⊕ (xn−1 · yn−1 )

where · denotes the AND operation, i.e.

0 · 0 = 0, 0 · 1 = 0, 1 · 0 = 0, 1 · 1 = 1.

What is the minimum number of bits Bob has to send to Alice so that she
can compute this function?

Solution 1. (i) Obviously Bob has to send only one bit, the one he finds
by computing y0 ⊕ y1 ⊕ · · · ⊕ yn−1 .
(ii) Bob must send all n bits in order for Alice to compute f .

Problem 2. Find all xA , xB , xC ∈ {0, 1} such that xA + xB + xC =

1 mod 2. We use the mapping f1 : {0, 1} → U (2)

f1 (0) := UH , f1 (1) := I2

where UH is the Walsh-Hadamard transform and U (2) denotes the unitary

group over C2 . Thus we can map from the triple (xA , xB , xC ) to linear
operators acting on three qubits

f3 (xA , xB , xC ) := f1 (xA ) ⊗ f1 (xB ) ⊗ f1 (xC ).

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 328

328 Problems and Solutions

Let
1
|ψi := (|001i + |010i + |100i − |111i).
2
For each triple (xA , xB , xC ) found in the first part of the problem, calculate

|φi := f3 (xA , xB , xC )|ψi.

Let sA , sB , sC denote the result (0 or 1) of measuring the first, second and

third qubit, respectively of |φi in the computational basis. In each case
determine
sA + sB + sC mod 2, xA · xB · xC .

Solution 2. We have

(xA , xB , xC ) ∈ { (0, 0, 1), (0, 1, 0), (1, 0, 0), (1, 1, 1) }.

We note the symmetry of the state |ψi with respect to qubit ordering. Thus
we need only to calculate the transform for (0, 0, 1) and (1, 1, 1). For (1, 1, 1)
we have f3 (1, 1, 1)|ψi = (I2 ⊗ I2 ⊗ I2 )|ψi = |ψi. Measuring the qubits yields

(sA , sB , sC ) ∈ { (0, 0, 1), (0, 1, 0), (1, 0, 0), (1, 1, 1) }

with equal probability. In each case we find sA + sB + sC = 1 mod 2. For

(0, 0, 1) we have f3 (0, 0, 1) = UH ⊗ UH ⊗ I2 . Since

1 1
|ψi = (|01i + |10i) ⊗ |0i + (|00i − |11i) ⊗ |1i
2 2
we obtain
1 1
f3 (0, 0, 1)|ψi = (|00i − |11i) ⊗ |0i + (|01i + |10i) ⊗ |1i.
2 2
We find that measuring the qubits yields

(sA , sB , sC ) ∈ { (0, 0, 0), (0, 1, 1), (1, 0, 1), (1, 1, 0) }

with equal probability. In each case we find sA + sB + sC = 0 mod 2.

(xA , xB , xC ) xA · xB · xC sA + sB + sC mod 2
(0,0,1) 0 0
(0,1,0) 0 0
(1,0,0) 0 0
(1,1,1) 1 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 329

Quantum Algorithms 329

We find that sA + sB + sC = xA · xB · xC mod 2. Suppose Alice, Bob

and Carol each have a bit string (xA,1 , . . . , xA,n ), (xB,1 , . . . , xB,n ) and
(xC,1 , . . . , xC,n ), respectively. They want to calculate
n
X
f (xA , xB , xC ) = (xA,j · xB,j · xC,j ) mod 2
j=1

sharing (communicating) as little information as possible. If Alice, Bob

and Carol share n triplets of qubits in the state |ψi they can calculate
sA,1 , . . . , sA,n , sB,1 , . . . , sB,n and sC,1 , . . . , sC,n respectively as above. Thus
n
X
f (xA , xB , xC ) = (sA,j + sB,j + sC,j ) mod 2.
j=1

If Alice, Bob and Carol calculate

n
X
SA|B|C = SA|B|C,j mod 2.
j=1

Bob and Carol need only to send one bit each (SB and SC ) to Alice for
Alice to compute f (xA , xB , xC ) = SA + SB + SC , for any n. In other words
the communication complexity is 2. Classically, for n ≥ 3, three bits of
communication are required.

Problem 3. (i) Find all x, y, z ∈ {0, 1, 2, 3} such that

x + y + z = 0 mod 2. (1)

What are the possible values of the function

(x + y + z) mod 4
f (x, y, z) :=
2
when the condition (1) holds?
(ii) Now use the binary representation for x = x1 x0 , y = y1 y0 and z = z1 z0
where x0 , x1 , y0 , y1 , z0 , z1 ∈ {0, 1}. Describe the condition x + y + z =
0 mod 2 in terms of x0 , x1 , y0 , y1 , z0 and z1 .
(iii) We use the map

f1 (0) = I2 , f1 (1) = UH .

Thus we can map from the triple (x0 , y0 , z0 ) to linear operators acting on
three qubits
f3 (x0 , y0 , z0 ) = f1 (x0 ) ⊗ f1 (y0 ) ⊗ f1 (z0 ).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 330

330 Problems and Solutions

Let
1
|ψi := (|000i − |011i − |101i − |110i).
2
For each triple (x0 , y0 , z0 ) found in part (i) calculate

|φi := f3 (x0 , y0 , z0 )|ψi.

Let sx , sy , sz denote the result (0 or 1) of measuring the first, second and

third qubit, respectively of |φi in the computational basis. In each case
determine
sx + sy + sz mod 2, x0 + y0 + z0 .

Solution 3. (i) Obviously x + y + z must be even. Thus the sum includes

only an even number (0 or 2) of odd numbers. Thus we have the nine
combinations

(0, 0, 0), (0, 1, 1), (0, 0, 2), (1, 1, 2), (0, 2, 2), (0, 1, 3), (2, 2, 2), (1, 2, 3), (0, 3, 3).

(ii) Let (x, y, z) be an element of the set of all permutations of elements of

the above set. When x + y + z is even, (x + y + z) mod 4 ∈ {0, 2}. Now
when x+y +z = 0 mod 2 then f (x, y, z) ∈ {0, 1}. Since x+y +z = 0 mod 2
the least significant bit of the sum must be zero. The least significant bit
is given by x0 ⊕ y0 ⊕ z0 = 0. We find that

f (x, y, z) = x1 ⊕ y1 ⊕ z1 ⊕ (x0 + y0 + z0 ).

XOR is denoted by “⊕” and OR is denoted by “+”. Thus we have

(x0 , y0 , z0 ) ∈ { (0, 0, 0), (0, 1, 1), (1, 0, 1), (1, 1, 0) }.

(iii) We note the symmetry of the state |ψi with respect to the qubit or-
dering. Thus we need only calculate the transform for (0, 0, 0) and (0, 1, 1).
For (0, 0, 0) we have

f3 (0, 0, 0)|ψi = I2 ⊗ I2 ⊗ I2 |ψi = |ψi.

Measuring the qubits yields

(sx , sy , sz ) ∈ { (0, 0, 0), (0, 1, 1), (1, 0, 1), (1, 1, 0) }

with equal probability. In each case we find sx + sy + sz = 0 mod 2. For

(0, 1, 1) we have f3 (0, 1, 1) = I2 ⊗ UH ⊗ UH . Note that |ψi can be written
as
1 1
|ψi = |0i ⊗ (|00i − |11i) − |1i ⊗ (|01i + |10i).
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 331

Quantum Algorithms 331

Therefore
1 1
f3 (0, 1, 1)|ψi = |0i ⊗ (|01i + |10i) − |1i ⊗ (|00i − |11i).
2 2
We find that measuring the qubits yields
(sx , sy , sz ) ∈ { (0, 1, 0), (1, 0, 0), (0, 0, 1), (1, 1, 1) }
with equal probability. In each case sx + sy + sz = 1 mod 2.
(x0 , y0 , z0 ) x0 + y0 + z0 sx + sy + sz mod 2
(0,0,0) 0 0
(0,1,1) 1 1
(1,0,1) 1 1
(1,1,0) 1 1
We find that (sx + sy + sz mod 2) = x0 + y0 + z0 . Thus for three parties
to calculate f (x, y, z), where each party has one of the x, y and z, it is
sufficient for each party to send one bit (x1 ⊕ sx or y1 ⊕ sy or z1 ⊕ sz ) to the
other parties to calculate f (x, y, z). In other words each party can calculate
x1 ⊕ sx ⊕ y1 ⊕ sy ⊕ z1 ⊕ sz = x1 ⊕ y1 ⊕ z1 ⊕ (x0 + y0 + z0 ) = f (x, y, z)
after communication. In other words three bits broadcast to all parties
are sufficient to calculate f (x, y, z), the communication complexity is 3 bits.
Classically it is necessary that 4 bits be broadcast.

Problem 4. (i) Determine the eigenvalues and eigenvectors of

A(x) := (1 − x)I2 + xUN OT , x ∈ {0, 1}.
(ii) Show that the unitary transform
Uf = |0f (0)ih00| + |0f (0)ih01| + |1f (1)ih10| + |1f (1)ih11|
where f : {0, 1} → {0, 1} is a boolean function and x denotes the boolean
negation of x, can be written as
Uf = |0ih0| ⊗ A(f (0)) + |1ih1| ⊗ A(f (1)).
(iii) Calculate  
1
Uf I2 ⊗ √ (|0i − |1i) .
2
Consider the cases f (0) = f (1) and f (0) 6= f (1).

Solution 4. (i) We have A(0) = I2 and A(1) = UN OT . Thus A0 has

eigenvalues 1 (twice), and A(1) has eigenvalues 1 and −1. We tabulate the
eigenvalues and corresponding eigenvectors of A(x)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 332

332 Problems and Solutions

eigenvalue eigenvector
1 √1 (|0i + |1i)
2
(−1)x √1 (|0i − |1i)
2

(ii) We have
Uf = (1 − f (0))|00ih00| + f (0)|01ih00|
+f (0)|00ih01| + (1 − f (0))|01ih01|
+(1 − f (1))|10ih10| + f (1)|11ih10|
+f (1)|10ih11| + (1 − f (1))|11ih11|
= |0ih0| ⊗ ((1 − f (0))|0ih0| + f (0)|1ih0|)
+|0ih0| ⊗ (f (0)|0ih1| + (1 − f (0))|1ih1|)
+|1ih1| ⊗ ((1 − f (1))|0ih0| + f (1)|1ih0|)
+|1ih1| ⊗ (f (1)|0ih1| + (1 − f (1))|1ih1|)
= |0ih0| ⊗ ((1 − f (0))(|0ih0| + |1ih1|) + f (0)(|0ih1| + |1ih0|))
+|1ih1| ⊗ ((1 − f (1))(|0ih0| + |1ih1|) + f (1)(|0ih1| + |1ih0|))
= |0ih0| ⊗ A(f (0)) + |1ih1| ⊗ A(f (1)).
(iii) We find
 
1 1
Uf I2 ⊗ √ (|0i − |1i) = |0ih0| ⊗ A(f (0)) √ (|0i − |1i)
2 2
1
+|1ih1| ⊗ A(f (1)) √ (|0i − |1i)
2
1
= |0ih0| ⊗ (−1)f (0) √ (|0i − |1i)
2
f (1) 1
+|1ih1| ⊗ (−1) √ (|0i − |1i)
2
1
= (−1)f (0) |0ih0| ⊗ √ (|0i − |1i)
2
f (1) 1
+(−1) |1ih1| ⊗ √ (|0i − |1i)
2
   
= (−1)f (0) |0ih0| + (−1)f (0)+f (1) |1ih1| ⊗ I2
 
1
× I2 ⊗ √ (|0i − |1i) .
2
Thus when f (0) = f (1) we apply the identity operator to the first qubit
and when f (0) 6= f (1) we apply a phase change to the first qubit. The
eigenvalues (−1)f (0) and (−1)f (1) are said to kick back to the first qubit.
A phase change combined with two Walsh-Hadamard transforms in the
appropriate order implements a NOT gate.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 333

Quantum Algorithms 333

Problem 5. (i) Alice and Bob share n entangled pairs of the form
√1 (|00i + |11i). We can write their shared state of 2n qubits in the form
2
of the generalized Bell state
n
2 −1
1 X
|ψi = √ |ji ⊗ |ji (1)
2n j=0

where the first n qubits belong to Alice and the second n qubits belong
to Bob. Furthermore Alice has 2n bits a0 , . . . , a2n −1 and Bob has 2n bits
b0 , . . . , b2n −1 . Let the unitary operators UP A and UP B act on the compu-
tational basis as follows

UP A |ji = (−1)aj |ji, j = 0, 1, . . . , 2n − 1

UP B |ji = (−1)bj |ji, j = 0, 1, . . . , 2n − 1.

Let
|φi := (UP A ⊗ UP B )|ψi. (2)
Calculate the state
! !
O O
UH ⊗ UH |φi. (3)
n n

(ii) For each of the cases

(a) a0 = b0 , a1 = b1 , . . . , a2n −1 = b2n −1

n
2X −1
(b) |ak − bk | = 2n−1
k=0

determine when measurement of the first n qubits in the computational

basis yields the same result as measurement of the second n qubits in the
computational basis.

Solution 5. (i) From (1) and (2) we obtain

n
2 −1
1 X
|φi = √ (−1)aj +bj |ji ⊗ |ji.
2n j=0

Thus we find for (3)

n n n
−1 2X
−1 2X−1
! ! 2X
O O 1
UH ⊗ UH |φi = √ (−1)aj +bj +j∗k+j∗l |ki⊗|li
n n
(2 2)n j=0 k=0 l=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 334

334 Problems and Solutions

since
! n−1
O O
UH |ji = UH |js i
n s=0
n−1
O 1
= UH √ (|0i + (−1)js |1i)
s=0
2
n
2X −1
= (−1)j0 k0 +j1 k1 +···+jn−1 kn−1 |ki
k=0

where we decompose j and k as follows

j = j0 + j1 2 + j2 4 + · · · + jn−1 2n−1 , k = k0 + k1 2 + k2 4 + · · · + kn−1 2n−1

and

j ∗ k := (j0 · k0 ) ⊕ (j1 · k1 ) ⊕ · · · ⊕ (jn−1 · kn−1 )

= j0 k0 + j1 k1 + · · · + jn−1 kn−1 mod 2.

(ii) For the case (a) we have for k = l

n n
−1
2X −1
2X
1 1
√ (−1)aj +bj +j∗k+j∗l = √ (−1)j∗k+j∗l
(2 2)n j=0 (2 2)n j=0
n
−1
2X
1
= √ (−1)j∗(k+l)
(2 2)n j=0
1
= √ 2n
(2 2)n
= 2−n/2 .

In other words the probability of measuring |ki ⊗ |ki for a given k is 2−n .
Furthermore n n
2X −1 2X −1
−n −n
2 =2 1 = 1.
k=0 k=0

For the case (b) we find when k = l

n n
2X −1 2X −1
aj +bj +j∗k+j∗l
(−1) = (−1)aj +bj = 0 .
j=0 j=0

Thus if condition (a) holds measuring the 2n qubits in the computational

basis always yields |ji and |ji, i.e. the first n qubits always yield exactly
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 335

Quantum Algorithms 335

the same result as the second n qubits.

If condition (b) holds then measuring the 2n qubits in the computational

basis yields |ji and |ki where j 6= k, i.e. the first n qubits never yield the
same result as the second n qubits.

Problem 6. (i) Show that the vectors

1 1
|0H i := √ (|0i + |1i), |1H i := √ (|0i − |1i)
2 2

form an orthonormal basis for C2 .

(ii) Determine the probabilities associated with finding |0i in the states
|0H i and |1H i.
(iii) Determine how to obtain |0H i and |1H i using only measurement and
the phase change operation

UP S := |0ih0| − |1ih1|.

(iv) Let f : {0, 1} → {0, 1} be a boolean function and

Uf := |0f (0)ih00| + |0f (0)ih01| + |1f (1)ih10| + |1f (1)ih11|.

Determine in terms of |0H i and |1H i

(a) Uf (|0H i ⊗ |0H i), (b) Uf (|0H i ⊗ |1H i).

These techniques are used to solve Deutsch’s problem.

Solution 6. (i) First we demonstrate the linear independence of the

vectors
1 1
a|0i + b|1i = √ (a + b)|0H i + √ (a − b)|1H i
2 2
1 1
a|0H i + b|1H i = √ (a + b)|0i + √ (a − b)|1i.
2 2
Thus for a|0H i + b|1H i = 0 it follows that a = b = 0.
(ii) We find

1
h0H |0H i = (h0|0i + h0|1i + h1|0i + h1|1i) = 1
2
1
h1H |1H i = (h0|0i − h0|1i − h1|0i + h1|1i) = 1
2
1
h0H |1H i = (h0|0i − h0|1i + h1|0i − h1|1i) = 0
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 336

336 Problems and Solutions

1 1
|h0|0H i|2 = |(h0|0i + h0|1i|2 =
2 2
1 1
|h0|1H i|2 = |(h0|0i − h0|1i|2 = .
2 2
Thus measurement projects the state |0i onto |0H i and |1H i with equal
probability.
(iii) Starting with |0i, we can obtain |0H i and |1H i by measurement in the
|0H i and |1H i basis and applying UP S as follows

Desired state Measure Transform

|0H i |0H i I2
|0H i |1H i UP S
|1H i |0H i UP S
|1H i |1H i I2

(iv) For (a) we have

1
|0H i ⊗ |0H i = (|00i + |01i + |10i + |11i).
2
Thus
1
Uf |0H i ⊗ |0H i = (|0f (0)i + |0f (0)i + |1f (1)i + |1f (1)i)
2
1
= (1 − f (0))|00i + f (0)|01i + f (0)|00i + (1 − f (0))|01i
2 
+(1 − f (1))|10i + f (1)|11i + f (1)|10i + (1 − f (1))|11i
= |0H i ⊗ |0H i.

For (b) we have

1
|0H i ⊗ |1H i = (|00i − |01i + |10i − |11i).
2
Thus
1
Uf |0H i ⊗ |1H i = (|0f (0)i − |0f (0)i + |1f (1)i − |1f (1)i)
2
1
= ((1 − f (0))|00i + f (0)|01i − f (0)|00i − (1 − f (0))|01i
2
+(1 − f (1))|10i + f (1)|11i − f (1)|10i − (1 − f (1))|11i)
1
= ((1 − 2f (0))|00i − (1 − 2f (0))|01i
2
+(1 − 2f (1))|10i − (1 − 2f (1))|11i)
1
= ((−1)f (0) |00i − (−1)f (0) |01i
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 337

Quantum Algorithms 337

−(−1)f (1) |10i − (−1)f (1) |11i)

1 
= (−1)f (0) |0i ⊗ (|0i − |1i) + (−1)f (1) |1i(|0i − |1i)
2
1
= √ (−1)f (0) ((|0i − (−1)f (0)+f (1) |1i) ⊗ |1H i)
2
1
= √ (−1)f (0) ((|0i − (−1)f (0)⊕f (1) |1i) ⊗ |1H i)
2
= (−1)f (0) |f (0) ⊕ f (1)H i ⊗ |1H i.
Note that f (0) ⊕ f (1) is 0 when f is constant, and 1 when f is balanced.
Thus by determining f (0) ⊕ f (1) we have solved Deutsch’s problem .

Problem 7. Consider the following quantum game Gn with n ≥ 3 players.

Each player Pj (j = 0, 1, . . . , n − 1) receives a single input bit xj and has
to produce a single output bit yj . It is known that there is an even number
of 1s among the inputs. The players are not allowed to communicate after
receiving their inputs. Then they are challenged to produce a collective
output that contains an even number of 1s if and only if the number of 1s
in the input is divisible by 4. Therefore, we require that
n−1 n−1
X 1X
yj ≡ xj (mod) 2
j=0
2 j=0

provided that
n−1
X
xj ≡ 0 (mod) 2.
j=0

We call x = x0 x1 · · · xn−1 the question and y = y0 y1 · · · yn−1 the answer.

Show that if the n-players are allowed to share prior entanglement, then
they can always win the game Gn .

Solution 7. We define the following n-qubit entangled state in the Hilbert

n
space C2
1
|ψ+ i := √ (|00 · · · 0i + |11 · · · 1i)
2
1
|ψ− i := √ (|00 · · · 0i − |11 · · · 1i).
2
The Walsh-Hadamard transform is given by
1 1 1 1
UH |0i → √ |0i + √ |1i, UH |1i → √ |0i − √ |1i.
2 2 2 2
Furthermore consider the unitary transformation
US |0i → |0i, US |1i → eiπ/2 |1i
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 338

338 Problems and Solutions

where eiπ/2 = i. If the unitary transformation US is applied to any two

qubits of |ψ+ i, while other qubits are left undisturbed, then

US |ψ+ i = |ψ− i

and if US is applied to any two qubits of |ψ− i, then

US |ψ− i = |ψ+ i.

Therefore, if the qubits of |ψ+ i are distributed among n players, and if

exactly m of them apply S to their qubit, the resulting state will be |ψ+ i
if m ≡ 0 (mod 4) and |ψ− i if m ≡ 2 (mod 4). The effect of applying the
Walsh-Hadamard transform to each qubit in |ψ+ i is to produce an equal
superposition of all classical n-bit strings that contain an even number of
1s, whereas the effect of applying the Walsh-Hadamard transform to each
qubit in |ψ− i is to produce an equal superposition of all classical n-bits
that contain an odd number of 1s. Thus
1 X
(UH ⊗ UH ⊗ · · · ⊗ UH )|ψ+ i = √ |y0 y1 · · · yn−1 i
2n−1 ∆(y)=0 (mod 2)

1 X
(UH ⊗ UH ⊗ · · · ⊗ UH )|ψ− i = √ |y0 y1 · · · yn−1 i
2n−1 ∆(y)=1 (mod 2)

where
n−1
X
∆(y) := yj
j=0

denotes the Hamming weight of y. Consequently the strategy is as follows:

At the beginning the state |ψ+ i is produced and its n-qubits are distributed
among the n players. After the separation each player Aj receives input
bit xj and does the following

1. If xj = 1, Aj applies the unitary transformation US to his qubit; other-

wise he/she does nothing.

2. He/she applies UH to this qubit.

3. He/she measures his/her qubit in order to obtain yj .

4. He/she produces yj as his/her output.

An even number of players will apply

Pn−1US to their qubit. If that number is
divisible by 4, which means that 21 j=0 xj is even, then the states reverts
to |ψ+ i after step 1 and therefore to a superposition of all |y0 y1 · · · yn−1 i
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 339

Quantum Algorithms 339

Pn−1
such that ∆(y) ≡ 0 (mod 2) after step 2. It follows that j=0 yi , the
number of players who measure and output 1, is even. If the number
of players who Papply S to their qubit is congruent to 2 modulo 4, which
n−1
means that 21 j=0 xj is odd, then the state evolves to |ψ− i after step 1 and
therefore to a superposition of all |yi ≡ |y0 y1 · · · yn−1 i such that ∆(y) ≡ 1
Pn−1
(mod 2) after step 2. In this case j=0 yj is odd. In either case, (1) is
satisfied at the end of the protocol.

Problem 8. Let x0 , x1 , y0 , y1 ∈ {0, 1} where Alice has x0 and x1 and Bob

has y0 and y1 . Alice and Bob want to calculate the boolean function

g(x0 , x1 , y0 , y1 ) := x1 ⊕ y1 ⊕ (x0 · y0 )

where ⊕ denotes the XOR operation and · denotes the AND operation.
Furthermore Alice and Bob share an EPR pair (Bell state)
1
√ (|00i − |11i).
2
Alice applies the unitary matrix
 π π
U R − + x0 ⊗ I2
16 4
to her qubit of the EPR pair and Bob applies the unitary matrix
 π π
I2 ⊗ UR − + y0
16 4
to his qubit of the EPR pair, where
 
cos(θ) − sin(θ)
UR (θ) := .
sin(θ) cos(θ)

Let a denote the result of Alice measuring her qubit of the EPR pair and
let b denote the result of Bob measuring his qubit of the EPR pair. Find
the probability that a ⊕ b = x0 · y0 , where ⊕ denotes the boolean XOR
operation and · denotes the boolean AND operation.

Solution 8. We define |ψi to be the state of the EPR-pair after Alice

and Bob apply their transforms. Consequently
 π π  π π 1
|ψi := UR − + x0 ⊗ U R − + y0 √ (|00i − |11i).
16 4 16 4 2
Thus
1   π π  π π 
|ψi = √ cos − + x0 |0i + sin − + x0 |1i
2 16 4 16 4
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 340

340 Problems and Solutions

  π π  π π 
⊗ cos − + y0 |0i + sin − + y0 |1i
  16 π 4 
π 16
π
4  
π
− − sin − + x0 |0i + cos − + x0 |1i
 π 16 4 16 4 
 π   π π
⊗ − sin − + y0 |0i + cos − + y0 |1i
16 4 16 4
1   π π   π π
= √ cos − + (x0 + y0 ) |00i + sin − + (x0 + y0 ) |01i
2 8 4 8 4
 π π   π π  
+ sin − + (x0 + y0 ) |10i − cos − + (x0 + y0 ) |11i .
8 4 8 4
Thus we find for the probabilities of obtaining a and b
a b a⊕b P (a, b)
1 2
0 0 0 2 cos (−π/8 + (x0 + y0 )π/4)
1 2
0 1 1 2 sin (−π/8 + (x0 + y0 )π/4)
1 2
1 0 1 2 sin (−π/8 + (x0 + y0 )π/4)
1 2
1 1 0 2 cos (−π/8 + (x0 + y0 )π/4)

Next we find the probability that

a ⊕ b = x0 · y0

for given x0 and y0

x0 y0 x0 · y0 P (a ⊕ b = x0 · y0 )
0 0 0 P (a = 0, b = 0) + P (a = 1, b = 1) = cos2 (π/8)
0 1 0 P (a = 0, b = 0) + P (a = 1, b = 1) = cos2 (π/8)
1 0 0 P (a = 0, b = 0) + P (a = 1, b = 1) = cos2 (π/8)
1 1 1 P (a = 1, b = 0) + P (a = 0, b = 1) = cos2 (π/8)
We find the probability

P (a ⊕ b = x0 · y0 ) = cos2 (π/8).

Problem 9. Let G be a finite group and let ρ : G → GL(n, C) be a

representation of G where n ∈ N. Show that
X
C := ρ(g)∗ ρ(g)
g∈G

√ √ −1
is positive definite. Thus the positive definite matrices C −1 , C and C
exist. Is ρC : G → GL(n, C) defined by
√ √ −1
ρC (g) := Cρ(g) C
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 341

Quantum Algorithms 341

a unitary representation of G?

Solution 9. Clearly C is hermitian. Since ρ(g) is invertible, we have

ρ(g)x 6= 0 for all x 6= 0, x ∈ Cn . Thus
X X
x∗ Cx = x∗ ρ(g)∗ ρ(g)x = kρ(g)xk2 > 0.
g∈G g∈G

It follows that C is positive definite and the positive definite matrices C −1 ,

√ √ −1
C and C exist. Obviously ρC (g) is invertible, then the inverse is
√ √ −1
given by ρC (g)−1 = Cρ(g)−1 C . Let g1 , g2 ∈ G then
√ √ −1 √ √ −1 √ √ −1
ρC (g1 )ρC (g2 ) = Cρ(g1 ) C Cρ(g2 ) C = Cρ(g1 )ρ(g2 ) C
√ √ −1
= Cρ(g1 · g2 ) C
where · is the group operation. Thus ρC provides a representation for G.
We also have for all g 0 ∈ G
X X
ρ(g 0 )∗ Cρ(g 0 ) = ρ(g 0 )∗ ρ(g)∗ ρ(g)ρ(g 0 ) = (ρ(g)ρ(g 0 ))∗ ρ(g)ρ(g 0 )
g∈G g∈G
X X
= (ρ(g · g 0 ))∗ ρ(g · g 0 ) = ρ(g)∗ ρ(g) = C
g∈G g∈G
0
where we used that {g · g : g ∈ G} = G. Now we find
√ −1 ∗ √ ∗√ √ −1 √ −1 √ −1
ρC (g)∗ ρC (g) = C ρ(g)∗ C Cρ(g) C = C ρ(g)∗ Cρ(g) C
√ −1 √ −1
= C C C = In
where In is the identity operator on Cn . Thus ρC is a unitary representation
for G.

Problem 10. Let G be a finite abelian group with identity 0 and let
f : G → S for some finite set S. We use the orthonormal bases
{ |giG : g ∈ G } ⊂ C|G| and { |siS : s ∈ S } ⊂ C|S| .
Find a unitary operator Uf such that
Uf (|giG ⊗ |s0 iS ) = |giG ⊗ |f (g)iS
where s0 ∈ S.

Solution 10. Let hg : S \ {s0 } → S \ {f (g)} be arbitrary one to one

functions for each g ∈ G. One solution is
 
X X
Uf = |giG hg|G ⊗ |f (g)iS hs0 |S + |hg (s)iS hs|S  .
g∈G s∈S\{s0 }
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 342

342 Problems and Solutions

Problem 11. Let G be a finite abelian group with identity 0 and let
f : G → S for some finite set S. We use the orthonormal bases

{ |gi : g ∈ G } ⊂ C|G| .

Discuss the quantum Fourier transform over the group G.

Solution 11. Since G is a finite abelian group the irreducible repre-

sentations of G are of the form χ : G → C. There are |G| such repre-
sentations labeled χg for g ∈ G. Since χ is a group homomorphism we
have χ(g1 · g2 ) = χ(g1 )χ(g2 ) for all g1 , g2 ∈ G. Irreducible representations
χ1 6= χ2 have the property that
X
χ1 (g)χ2 (g) = 0.
g∈G

The trivial irreducible representation χ1 (g) = 1 for all g ∈ G provides

X X
χ(g)χ1 (g) = χ(g) = 0
g∈G g∈G

for nontrivial χ. The transform over the group structure is

1 XX
UF (G) := p χu (v)|uihv|
|G| u∈G v∈G

where the χl : G → C are the |G| irreducible representations of G, χl (g) is

a |G|-th root of unity. For G = ZN , the group of integers {0, 1, . . . , N − 1}
with addition modulo N , we obtain the quantum Fourier transform.

Problem 12. Let G be a finite abelian group with identity 0, and K

be a subgroup of G. The cosets of K, i.e. g · K := { g · k : k ∈ K } for
g ∈ G, partition G. Now suppose f : G → X, for some finite set X, with
the property f (g1 ) = f (g2 ) if and only if g1 · K = g2 · K. From f and G we
wish to find the hidden subgroup K. As the computational basis we use
the orthonormal basis BG = { |giG : g ∈ G } ⊂ C|G| and the orthonormal
basis BX = { |xiX : x ∈ X } ⊂ C|X| . We begin with the state

1 X
p |giG ⊗ |f (g)iX .
|G| g∈G

Show how measurement of the second system in the basis BX and then
performing the quantum Fourier transform over the group structure for the
first system can be used to find K.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 343

Quantum Algorithms 343

Solution 12. We can factor out the second state according to the cosets
of K !
1 X X
p |g · kiG ⊗ |f (g)iX .
|G| g·K k∈K
Measuring the second state in the basis BX projects the first state, for some
g, onto
1 X
p |g · kiG .
|K| k∈K
Applying UF (G) yields
!
1 X 1 X X
UF (G) p |g · kiG = p χu (g · k) |uiG
|K| k∈K |G| · |K| u∈G k∈K
!
1 X X
=p χu (g) χu (k) |uiG
|G| · |K| u∈G k∈K

since the irreducible representations are group homomorphisms. Due to the

fact that the irreducible representation χu of G can also be considered as
an irreducible representation of the subgroup K we have
X
χ(k) = 0
k∈K

for nontrivial χ so that

!
1 X 1 X X
UF (G) p |g · kiG = p χu (g) χu (k) |uiG
|K| k∈K |G| · |K| u∈G k∈K
s
|K| X
= χu (g)|uiG .
|G| u∈G
χu (k)=1 ∀k∈K

In the last result we find non-zero probability amplitudes for |uiG when
χu (k) = 1 for all k ∈ K. Measurement of the state yields u ∈ G with
χu (k) = 1 ∀k ∈ K. We can consequently test the different K against χu . If
K has a generator k0 , then χu (k0 ) = 1 so that measurement provides the
possibility of recovering the generator of K. In general, the process must
be repeated since we may have found a generator for a subgroup.

Problem 13. (i) The CHSH game is a game between a referee from one
side and two player (named Alice and Bob) from the other side. Alice and
Bob are separated and not allowed to communicate till the game is over.
Let a, b, x, y ∈ {0, 1}. The game starts with a referee selecting two bits x
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 344

344 Problems and Solutions

and y uniformly at random. The referee then sends x to Alice and y to

Bob. Alice sends back to the referee a bit a and Bob sends back a bits
b. So Bob’s response bit b cannot depend on Alice’s input and vice versa
Alice’s response bit a cannot depend on Bob’s input. After the bits a and
b has been transferred to the referee. Then he tests whether the boolean
equation is satisfied
x · y = a ⊕ b.
If it is satisfied Alice and Bob win and the referee loses. If not the referee
wins and Alice and Bob lose. What is the best strategy for Alice and Bob
to win?
(ii) Let |0i, |1i be the standard basis in C2 , I2 be the 2 × 2 identity matrix,
   
cos(π/8) − sin(π/8) cos(π/8) sin(π/8)
RA = , RB =
sin(π/8) cos(π/8) − sin(π/8) cos(π/8)
and
UA = RA ⊗ I2 , UB = I2 ⊗ RB , UAB = RA ⊗ RB
where A refers to Alice and B refers to Bob. Note that RA RB = I2 .
Consider the Bell state
1
|ψi = √ (|0iA ⊗ |0iB + |1iA ⊗ |1iB ).
2
If Alice receives x = 0 and Bob receives y = 0, then calculate

|ψi
f = (I2 ⊗ I2 )|ψi, f 2.
p00 = |hψ||ψi|
If Alice receives x = 0 and Bob receives y = 1, then calculate

|ψi
f = (I2 ⊗ RB )|ψi, f 2.
p01 = |hψ||ψi|
If Alice receives x = 1 and Bob receives y = 0, then calculate

|ψi
f = (RA ⊗ I2 )|ψi, f 2.
p10 = |hψ||ψi|
If Alice receives x = 1 and Bob receives y = 1, then calculate

|ψi
f = (RA ⊗ RB )|ψi, f 2.
p11 = |hψ||ψi|
Calculate the probability
1 1 1 1
p= p00 + p01 + p10 + p11
4 4 4 4
and show that p > 3/4. Note that
√ √
q q
1 1
cos(π/8) = 2 + 2, sin(π/8) = 2 − 2,
2 2
1√
cos2 (π/8) − sin2 (π/8) = 2.
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 345

Quantum Algorithms 345

Solution 13. (i) The truth table is

x y a b x·y a⊕b w
0 0 0 0 0 0 1
0 0 0 1 0 1 0
0 0 1 0 0 1 0
0 0 1 1 0 0 1
0 1 0 0 0 0 1
0 1 0 1 0 1 0
0 1 1 0 0 1 0
0 1 1 1 0 0 1
1 0 0 0 0 0 1
1 0 0 1 0 1 0
1 0 1 0 0 1 0
1 0 1 1 0 0 1
1 1 0 0 1 0 0
1 1 0 1 1 1 1
1 1 1 0 1 1 1
1 1 1 1 1 0 0
From the truth table we find the following. If Alice always submits a = 0
and Bob always submits b = 0 the winning probability for Alice and Bob
is 3/4. Analogously, if Alice always submits a = 1 and Bob always submits
b = 1 the winning probability for Alice and Bob is 3/4.
(ii) Obviously for (x = 0, y = 0) we obtain p00 = 1. For (x = 0, y = 1) we
obtain p01 = cos2 (π/8). For (x = 1, y = 0) we obtain p10 = cos2 (π/8) For
(x = 1, y = 1) we have
1
(RA ⊗ RB )|ψi = (RA ⊗ RB ) √ (|0i ⊗ |0i + |1i ⊗ |1i)
2
1
= √ ((RA |0i) ⊗ (RB |0i) + (RA |1i) ⊗ (RB |1i))
2
1
= √ (cos(π/4)|0i ⊗ |0i − sin(π/4)|0i ⊗ |1i
2
+ sin(π/4)|1i ⊗ |0i + cos(π/4)|1i ⊗ |1i)
where we utilized that
cos2 (π/8)−sin2 (π/8) = cos(π/4) = sin(π/4), 2 sin(π/8) cos(π/8) = sin(π/4).
It follows that
1
|hψ|(RA ⊗ RB )|ψi|2 = .
2
Hence
√
1 3 1 2 5 2
p = (p00 + p01 + p10 + p11 ) = + cos (π/8) = + ≈ 0.8018
4 8 2 8 8
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 346

346 Problems and Solutions

Programming Problems

Problem 1. Consider two bitstrings b1, b2 of the same length n. We

define the scalar product of the two bitstrings as

b1 ? b2 = (b1[0] · b2[0]) ⊕ (b1[1] · · · b2[0]) ⊕ · ⊕ (b1[n − 1] · b2[n − 1])

where ⊕ is the XOR operation and · is the AND operation. Give a C++
implementation utilizing the bitset class.

Solution 1. The C++ program is

// scalarproduct.cpp
// c++ -std=c++11 -o scalarproduct scalarproduct.cpp
#include <iostream>
#include <string>
#include <bitset>
using namespace std;

int main(void)
{
// least significant bit on the right-hand side
const int n=8;
bitset<n> b1(string("10001010"));
bitset<n> b2(string("00110111"));
cout << b2[0] << endl << endl;
int temp[n];
int j;
for(j=0;j<n;j++)
{
temp[j] = b1[j] & b2[j];
cout << temp[j] << endl;
}
cout << endl;
int scalar = temp[0] ^ temp[1];
for(int j=2;j<n;j++)
{
scalar = scalar ^ temp[j];
}
cout << scalar << endl;
return 0;
}

Problem 2. Let a, b, x, y ∈ {0, 1}. Find all solutions of the boolean

equation
a · b = x ⊕ y.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 347

Quantum Algorithms 347

Utilize the bitset class of C++.

Solution 2. The C++ program provides the solutions

// bitequation.cpp
#include <iostream>
#include <bitset>
using namespace std;

int main(void)
{
bitset<1> a;
bitset<1> b;
bitset<1> x;
bitset<1> y;
int j1,j2,j3,j4;
cout << "The solutions are: " << endl;
for(j1=0;j1<=1;j1++)
for(j2=0;j2<=1;j2++)
for(j3=0;j3<=1;j3++)
for(j4=0;j4<=1;j4++)
{
a = j1; b = j2; x = j3; y = j4;
if((a & b)==(x ^ y))
{ cout << "a = " << j1 << " " << "b = " << j2 << " "
<< "x = " << j3 << " " << "y = " << j4 << endl;
}
}
return 0;
}

provides the solutions

The solutions are:

a = 0 b = 0 x = 0 y = 0
a = 0 b = 0 x = 1 y = 1
a = 0 b = 1 x = 0 y = 0
a = 0 b = 1 x = 1 y = 1
a = 1 b = 0 x = 0 y = 0
a = 1 b = 0 x = 1 y = 1
a = 1 b = 1 x = 0 y = 1
a = 1 b = 1 x = 1 y = 0

Problem 3. Let ω := e2πi/4 . The 4 × 4 quantum Fourier transform is

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 348

348 Problems and Solutions

given by
1 1 1 1
 
1 1 ω ω2 3
ω 
U=  .
2 1 ω2 ω4 ω6
1 ω3 ω6 ω 9

Since ω 4 = 1 we can write

1 1 1 1
 
1 1 ω ω2 3
ω 
U=  .
2 1 ω2 1 ω2
1 ω3 ω2 ω

Show that the matrix is unitary. Then find the four eigenvalues. Apply
matrix to the normalized vector

1
 
1 1
φ =  .
2 1
1

Discuss.

Solution 3. The Maxima program

/* Fourier.mac */
om: exp(2*%pi*%i/4);
F: matrix([1,1,1,1],[1,om,om^2,om^3],[1,om^2,1,om^2],[1,om^3,om^2,om])/2;
F: trigexpand(F);
FT: transpose(F);
FTC: conjugate(FT);
R: F . FTC;
phi: matrix([1],[1],[1],[1])/2;
psi: F . phi;
eigenvalues(F);

provides us with the information that F is unitary since F F ∗ = I4 . The

eigenvalues are +1 (twice), −1 and i and

1
 
0
Fφ =  .
0
0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 349

Quantum Algorithms 349

14.3 Supplementary Problems

Problem 1. Consider the standard basis

     
1 0 0
|1i =  0  , |2i =  1  , |3i =  0 
0 0 1

in C3 , the unitary matrix (quantum Fourier transform)

 
1 1 1
1 
U=√ 1 exp(i2π/3 exp(−i2π/3) 
3 1 exp(−i2π/3) exp(i2π/3)

and the six permutation matrices

     
1 0 0 0 1 0 0 0 1
P1 =  0 1 0  , P2 =  0 0 1  , P3 =  1 0 0,
0 0 1 1 0 0 0 1 0
     
0 0 1 0 1 0 1 0 0
P4 =  0 1 0  , P5 =  1 0 0  , P6 =  0 0 1.
1 0 0 0 0 1 0 1 0
The first three permutation matrices P1 , P2 , P3 are the even permutation
matrices with
det(P1 ) = det(P2 ) = det(P3 ) = 1
and the last three permutation matrices P4 , P5 , P6 are the odd permutation
matrices with
det(P4 ) = det(P5 ) = det(P6 ) = −1.
Do the following calculation. First find the normalized state

|ψ1 i = U |2i.

Then calculate the six normalized states

|ψk i = Pk |ψ1 i, k = 1, 2, 3, 4, 5, 6

Next find U ∗ and check it is really the inverse of U . Then calculate the six
normalized states
U ∗ |ψk i, k = 1, 2, 3, 4, 5, 6.
Finally find the twelve probabilities

|h2|U ∗ |ψk i|2 , |h3|U ∗ |ψk i|2 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 350

350 Problems and Solutions

Discuss. This exercise plays a role for the quantum permutation algorithm.

Problem 2. Suppose the input consists of n qubits and the boolean

function to be calculated is

f := { 0, 1, . . . , 2n − 1 } → { 0, 1}.

Show that n
2X −1 X
1
Uf := |jihj| ⊗ |k ⊕ f (j)ihk|
j=0 k=0

is a unitary matrix (a permutation matrix).

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 351

Chapter 15

Quantum Error
Correction

15.1 Introduction
In classical communication theory where bits are communicated, the only
possible type of error that can occur is a bit flip. In the quantum case
any rotation or phase change in the Hilbert space of the quantum state is
an error. Thus there are an infinite number of different errors that could
occur just for a single qubit. Fortunately the measurement process involves
the projection of the quantum state into a compatible subspace. Thus
measurement to determine the occurrence of an error reduces the error to
one compatible with the measurement. Suppose the data is contained in
the state |ψi, and the environment is described by the state |Ei. The
initial state of the entire system is described by the tensor product of the
states |ψi ⊗ |Ei, which evolves according to some unitary operation U . The
state |ψi evolves according to the unitary operation Uψ which describes a
quantum algorithm. In classical error correction codes, all that needs to be
corrected are bit flips. In the quantum case errors such as bit flips, phase
changes and rotations complicate the error correction techniques. Since
arbitrary errors in an encoding of information cannot be corrected, only
certain types of errors are assumed to occur. The types of errors depend on
the implementation. For example, suppose the types of errors (which we
assume are distinguishable due to an encoding) are described by the unitary

351
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 352

352 Problems and Solutions

basis E1 , . . . , En so that all errors are described by a linear combination

E = c1 E1 + · · · + cn En , Ej† Ej = I

where I is the identity operator and j = 1, . . . , n. We use the product

state |ψi ⊗ |0i, where |ψi is an encoded quantum state with the necessary
property that it can be used to determine if any error of E1 , . . . , En has
occurred, and the second quantum register will hold the number of the type
of error which occurred. Let S denote the operator for the error syndrome

S(Ej ⊗ I)|ψi ⊗ |0i := |ψi ⊗ |ji.

Now the encoded state with errors is given by

n
X
(E ⊗ I)(|ψi ⊗ |0i) = cj Ej |ψi ⊗ |0i.
j=1

Applying the operator for the error syndrome gives

n
X
S(E ⊗ I)(|ψi ⊗ |0i) = cj Ej |ψi ⊗ |ji.
j=1

Measuring the second register identifies the error. Suppose the measure-
ment corresponds to |ki, then the error can be repaired since

(Ek−1 ⊗ I)((Ek ⊗ I)|ψi ⊗ |ki) = |ψi ⊗ |ki.

Given a normalized state in C2

|ψi = α|0i + β|1i, |α|2 + |β|2 = 1.

A bit-flip changes the normalized state into

α|1i + β|0i

i.e. we apply the operator |0ih1|. The phase-flip error changes the normal-
ized state into
α|0i − β|1i
i.e. we apply the operator |0ih0| + eiπ |1ih1|.

Let σ1 , σ2 , σ3 be the Pauli spin matrices and σ0 = I2 . A general single bit

error is thus a map
e1 I2 + e2 σ1 + e3 σ2 + e4 σ3 .
Interaction with the environment maps single qubits as

|ψi 7→ (e1 σ0 + e2 σ1 + e3 σ2 + e4 σ3 )|ψi.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 353

Quantum Error Correction 353

15.2 Solved Problems

Problem 1. Calculate the following in terms of I2 , X, Y , Z

(i) XZ, ZX
(ii) UCN OT (X ⊗ I2 )UCN OT
(iii) UCN OT (I2 ⊗ X)UCN OT
(iv) UCN OT (Z ⊗ I2 )UCN OT
(v) UCN OT (I2 ⊗ Z)UCN OT
(vi) UCN OT (X ⊗ X)UCN OT
(vii) UCN OT (Z ⊗ Z)UCN OT
(viii) UCN OT UCN OT

where

I2 := |0ih0| + |1ih1|, X := |0ih1| + |1ih0|

Y := |0ih1| − |1ih0|, Z := |0ih0| − |1ih1|
UCN OT := |0ih0| ⊗ I2 + |1ih1| ⊗ X.

Solution 1. Straightforward calculation yields

(i) XZ = −Y, ZX = Y
(ii) UCN OT (X ⊗ I2 )UCN OT = X ⊗ X
(iii) UCN OT (I2 ⊗ X)UCN OT = I2 ⊗ X
(iv) UCN OT (Z ⊗ I2 )UCN OT = Z ⊗ I2
(v) UCN OT (I2 ⊗ Z)UCN OT = Z ⊗ Z
(vi) UCN OT (X ⊗ X)UCN OT = X ⊗ I2
(vii) UCN OT (Z ⊗ Z)UCN OT = I2 ⊗ Z
(viii) UCN OT UCN OT = I2 ⊗ I2 .

Problem 2. Suppose that the only errors which can occur to three qubits
are described by the set of 8 × 8 unitary matrices

{I2 ⊗I2 ⊗I2 , I2 ⊗UN OT ⊗UN OT , I2 ⊗UP ⊗UP , I2 ⊗(UP UN OT )⊗(UP UN OT )}

where UP := |0ih0| − |1ih1|, UN OT := |0ih1| + |1ih0|. A linear combination

of these unitary matrices is given by

E := αI2 ⊗ I2 ⊗ I2 + βI2 ⊗ UN OT ⊗ UN OT + δI2 ⊗ UP ⊗ UP

+γI2 ⊗ (UP UN OT ) ⊗ (UP UN OT )
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 354

354 Problems and Solutions

where α, β, δ, γ ∈ C. Describe how an arbitrary error E on the three-qubit

state
1
√ (|00i + |11i) ⊗ |ψi
2
can be corrected to obtain the correct |ψi as the last qubit, where

|ψi := a|0i + b|1i, |a|2 + |b|2 = 1, a, b ∈ C.

Solution 2. Applying the matrix (I8 ≡ I2 ⊗ I2 ⊗ I2 )

αI8 + βI2 ⊗ UN OT ⊗ UN OT + δI2 ⊗ UP ⊗ UP + γI2 ⊗ (UP UN OT ) ⊗ (UP UN OT )

to the state
1
√ (|00i + |11i) ⊗ |ψi
2
yields the state
1 1
α √ (|00i + |11i) ⊗ |ψi + β √ (|01i + |10i) ⊗ (a|1i + b|0i)
2 2
1 1
+δ √ (|00i − |11i) ⊗ (a|0i − b|1i) + γ √ (|01i − |10i) ⊗ (a|1i − b|0i).
2 2
Thus we measure the first two qubits in the Bell basis and apply the cor-
responding transform to the last qubit to obtain |ψi.

measure transform
√1 (|00i+ |11i) I2
2
√1 (|01i + |10i) UN OT
2
√1 (|00i − |11i) UP
2
√1 (|01i − |10i) UN OT UP
2

Problem 3. Assume that the only errors that occur in a system of

qubits are isolated to individual qubits, i.e. the error in one qubit state
is independent of the error in another qubit state. Hence the error for
each qubit can be expressed as a linear operator E on the Hilbert space
C2 . Furthermore E can be expressed as a linear combination of the 2 × 2
identity matrix and the Pauli spin matrices σ1 , σ2 and σ3 . Now consider
a non-degenerate n-qubit code representing a single qubit state which can
correct errors in up to k qubits.
(i) Find a lower bound describing n.
(ii) Find the lower bound for k = 1.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 355

Quantum Error Correction 355

Hint. The n-qubit states representing qubits with errors should be distinct
(orthogonal) for distinct errors and distinct from the case where there are
no errors.

Solution 3. (i) We have 3 distinct errors on a single qubit described by

the Pauli matrices. Thus there are
 
l n n!
3 ≡ 3l
l l!(n − l)!

distinct errors in l qubits of n qubits. The total number of ways to have at

most k errors in n qubits is then given by
k  
l n
X
3 .
l
l=0

There are 2n orthogonal states in a Hilbert space describing n qubits. Since

the states representing qubits (|0i or |1i) with distinct errors should be
orthogonal, we find
k  
l n
X
2 3 ≤ 2n .
l
l=0

(ii) For k = 1 we have the bound 2(1 + 3n) ≤ 2n . In other words, for k = 1
we find n ≥ 5.

Problem 4. Consider the Pauli matrices σ1 , σ2 , σ3 and the 2 × 2 unit

matrix. Do these matrices form a group under matrix multiplication? If
so, provide a proof. If not, what set with minimal cardinality includes the
Pauli matrices and forms a group under matrix multiplication?

Solution 4. Obviously these matrices do not form a group since σ1 σ2 =

iσ3 which is not a Pauli matrix. Since the factor i is introduced, all factors
which are powers of i must be included i.e. ±1 and ±i:

{±σ, ±iσ : σ ∈ {I2 , σ1 , σ2 , σ3 }}.

Thus we have a group of cardinality 16.

Problem 5. Let S be a set of operators closed under the hermitian

conjugate (adjoint) such that the quantum code CS of the set of states

CS := { |ψi : U |ψi = |ψi, ∀U ∈ S }

is non trivial (does not consist only of the zero state). The set S is called
the stabilizer of the code CS .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 356

356 Problems and Solutions

(i) Show that

[M, N ]|ψi = 0, ∀M, N ∈ S, |ψi ∈ CS .

(ii) Let [E, M ] = 0 for some M ∈ S. What can be said about E|ψi when
|ψi ∈ CS ?
(iii) Let [E, M ]+ = 0 for some M ∈ S. What can be said about E|ψi when
|ψi ∈ CS ?

Solution 5. (i) Since M N |ψi = M |ψi = |ψi, N M |ψi = N |ψi = |ψi it

follows that [M, N ]|ψi = 0.
(ii) For |ψi ∈ CS we have

M E|ψi = EM |ψi = E|ψi.

Thus E|ψi is an eigenstate of M corresponding to the eigenvalue 1.

(iii) For |φi, |ψi ∈ CS we have

hφ|E|ψi = hφ|EM |ψi = −hφ|M E|ψi = −hφ|M ∗ E|ψi = −hφ|E|ψi.

Thus hφ|E|ψi = 0. In other words, E|ψi ∈

/ CS . Furthermore

M E|ψi = −EM |ψi = −E|ψi

thus E|ψi is an eigenstate of M corresponding to the eigenvalue −1.

Problem 6. Consider the application of the controlled NOT gate to the

state |10i ≡ |1i ⊗ |0i. Suppose that a single qubit error σ1 ⊗ I2 occurs
before the controlled NOT operation is performed. How many errors does
the resulting state have after performing the control NOT operation? Also
discuss the case when the error is given by √12 (I2 + σ1 ) ⊗ I2 .

Solution 6. The intended operation is UCN OT |10i = |11i. The operation

with error is

UCN OT (σ1 ⊗ I2 )|10i = UCN OT |00i = |00i.

Thus we have two single qubit errors σ1 ⊗ I2 and I2 ⊗ σ1 . For the second
type of single qubit error we have
1 1 1
UCN OT √ (I2 + σ1 ) ⊗ I2 |10i = UCN OT √ (|10i + |00i) = √ (|00i + |11i).
2 2 2
Here we find that the error cannot be expressed in terms of single qubit
errors.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 357

Quantum Error Correction 357

Problem 7. Consider the application of the controlled NOT gate to

corresponding pairs of qubits of the state |1̃0̃i where |0̃i is a code word for
|0i and |1̃i is a code word for |1i in a quantum error correction code that
can correct single qubit errors. In other words the controlled NOT gate is
applied to the i-th qubit of |1̃i as control and the i-th qubit of |0̃i as target.
Suppose that a single qubit error
1
√ (I2 + σ1 )
2
on one of the qubits of |1̃i occurs before the controlled NOT operation is
performed. Discuss the error correction of this computation.

Solution 7. Suppose the single qubit error occurs on the j-th qubit, and
that the encoding |1̃i and |0̃i is n qubits long. Thus the error is
 
j−1
! 2n
1 O O
E := √ I2 ⊗ (I2 + σ1 ) ⊗  I2  .
2 k=1 k=j+1

Let ŨCN OT denote the pairwise application of the controlled NOT opera-
tion. Then
∗
ŨCN OT E|1̃0̃i = ŨCN OT E ŨCN OT ŨCN OT |1̃0̃i.

The controlled NOT operation is given by

1 1
UCN OT = (I2 + σ3 ) ⊗ I2 + (I2 − σ3 ) ⊗ σ1 .
2 2
We find the matrix
1 ∗ 1
√ UCN OT (I2 + σ1 )UCN OT = √ (I2 ⊗ I2 + σ1 ⊗ σ1 ).
2 2
Consequently we obtain the normalized state
2n
!
1 O
ŨCN OT E|1̃0̃i = √ I2
2 k=1
   
j−1 n+j−1
! 2n
1 O O O
+√ I2 ⊗ σ1 ⊗  I2  ⊗ σ1 ⊗  I2 
2 k=1 k=j+1 k=n+j+1

×ŨCN OT |1̃0̃i.

Thus there is a linear combination of no error, or a single qubit error in

the first block of n qubits and a single qubit error in the second block of n
qubits. Both errors can be independently corrected.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 358

358 Problems and Solutions

Programming Problem

Problem 1. Consider the single qubit state a|0i + b|1i with aa + bb = 1.

Encode the single qubit state in three qubits as

a(|0i ⊗ |0i ⊗ |0i) + b(|1i ⊗ |1i ⊗ |1i)

i.e. |0i 7→ |0i⊗|i⊗|0i, |1i 7→ |1i⊗|1i⊗|1i. Give a Maxima implementation.

Solution 1. Utilizing the UCN OT gate

 
0 1
UCN OT = I2 ⊕
1 0

and I2 ⊗ UCN OT , UCN OT ⊗ I2 we can implement the encoding.

/* encoding.mac */
v0: matrix([1],[0]); v1: matrix([0],[1]);
psi: a*v0 + b*v1;
alpha: kronecker_product(psi,kronecker_product(v0,v0));
I2: matrix([1,0],[0,1]);
UCNOT: matrix([1,0,0,0],[0,1,0,0],[0,0,0,1],[0,0,1,0]);
U1: kronecker_product(I2,UCNOT);
U2: kronecker_product(UCNOT,I2);
beta: U1 . U2 . alpha;

15.3 Supplementary Problem

Problem 1. Let |0i, |1i be the standard basis in C2 . Consider the

quantum bit
1
|ψi = √ (|0i − |1i)
2
which is encoded as
1
C|ψi = |φi = √ (|0i ⊗ |0i ⊗ |0i − |1i ⊗ |1i ⊗ |1i).
2

Consider the error E = 54 σ1 ⊗ I2 ⊗ I2 + 53 I2 ⊗ σ1 ⊗ I2 . Show that

4 3
E|φi = √ (|1i⊗|0i⊗|0i−|0i⊗|1i⊗|1i)+ √ (|0i⊗|1i⊗|0i−|1i⊗|0i⊗|1i).
5 2 5 2
Then apply the syndrome extraction to (E|ψi) ⊗ |0i ⊗ |0i ⊗ |0i.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 359

Chapter 16

Quantum Cryptography

16.1 Introduction
Cryptography usually involves a key or keys to be used in encryption and
decryption algorithms. Classical cryptography generally relies on maps that
are perceived to be very difficult to invert with incomplete information.
One popular algorithm due to Rivest, Shamir and Adelman is the RSA
algorithm

n = pq, M ∈ {0, 1, . . . , n − 1}, ed = 1 mod (p − 1)(q − 1)

C := (M e mod n), M ≡ (C d mod n)

where p and q are large prime numbers, M is the message and C is the
encrypted message. If p, q and d are unknown then, in general, C can-
not easily be obtained from M . However, Shor found that p and q can be
obtained with relative ease (and consequently also d) using the quantum
Fourier transform.

Quantum cryptography is concerned with the secure distribution of keys

using quantum communication channels. Another application is hiding
classical data in quantum states. Quantum cryptographic techniques rely
on physics to supply secure communication in the sense that it is possible to
determine if someone has intercepted the message. This is due to the fact
that measurement in quantum mechanics is associated with a disturbance
(projection) of the quantum state. Entanglement can also be used to detect

359
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 360

360 Problems and Solutions

whether a message has been intercepted. For example, when one qubit of
an EPR pair is measured the correlation is destroyed which can be tested
with Bell’s inequality.

16.2 Solved Problems

Problem 1. Let p, q be large prime numbers, (e, n) a public key, and
(d, n) a private key for the RSA cryptosystem where n = pq. Let M, C ∈
{0, 1, . . . , n − 1} with

C ≡ M e mod n, M ≡ C d mod n.

(i) Let gcd(a, n) = 1. Show that p and q can be determined from even
r ∈ N if ar ≡ 1 mod n, ar/2 6≡ 1 mod n, ar/2 6≡ 1 mod n.
(ii) Illustrate (i) with p = 5, q = 11 and a = 6.
(iii) Show that M can be determined from C and t ∈ N, where (t minimal)

C et ≡ C mod n.

(iv) Illustrate (iii) with p = 5, q = 11, e = 9 and C = 48.

Solution 1. (i) We have

r r
ar − 1 ≡ 0 mod n, (a 2 − 1)(a 2 + 1) ≡ 0 mod n

Consequently
r r
[(a 2 − 1) mod n][(a 2 + 1) mod n] = kn = kpq

for some k ∈ Z. Consequently one of p and q is given by

r
gcd(a 2 − 1 mod n, n).

The second is found by division of n. The gcd can be determined efficiently

using the Euclidean algorithm.
(ii) The powers (from 0) of a in modulo 5 · 11 = 55 arithmetic are

1, 6, 36, 216 ≡ 51, ≡ 306 ≡ 31, ≡ 186 ≡ 21, ≡ 126 ≡ 16, ≡ 96 ≡ 41,

≡ 246 ≡ 26, ≡ 156 ≡ 46, ≡ 276 ≡ 1

where the last power is 10. Thus we use r = 10. Thus the periodicity of
this sequence is also 10. Consequently

65 − 1 = 7775 ≡ 20 mod n.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 361

Quantum Cryptography 361

Using the Euclidean algorithm

55 = 2 · 20 + 15, 20 = 1 · 15 + 5, 15 = 3 · 5 + 0

Thus we find the gcd(20, 55) = 5. Consequently p = 5 and q = n/5 = 11.

(iii) We find

C t ≡ M et ≡ C det ≡ (C et )d ≡ C d ≡ M mod n.

(iv) We find
C e = 489 = 49 49 39 ≡ 14 · 14 · 48 ≡ 3.
The powers (from 0) of C e in modulo 5 · 11 = 55 arithmetic are

1, 3, 9, 27, 26, 23, 14, 42, 16, 48

i.e. we use t = 9. Now C t ≡ 3 ≡ M mod n.

Problem 2. Apply the quantum Fourier transform to the first register of

1
|ψi := √ (|0i ⊗ |0i + |1i ⊗ |1i + |2i ⊗ |2i + |3i ⊗ |0i + |4i ⊗ |1i + |5i ⊗ |2i)
6
to find the underlying periodicity in the second register. In this case the
quantum Fourier transform is given by
5
1 X −i2πjk/6
UQF T = √ e |jihk|.
6 j,k=0

Solution 2. Applying the quantum Fourier transform yields

5
1X 
(UQF T ⊗ I)|ψi = |ji ⊗ |0i + e−i2πj/6 |1i + e−i2πj2/6 |2i
6 j=0

+e−i2πj3/6 |0i + e−i2πj4/6 |1i + e−i2πj5/6 |2i
5
1 X  
1 + e−iπj |ji ⊗ |0i + e−iπj/3 |1i + e−i2πj/3 |2i

=
6 j=0
2
1X  
= |2ji ⊗ |0i + e−iπ2j/3 |1i + e−i4πj/3 |2i .
3 j=0

Thus measuring the first register yields 0, 2 or 4. The minimum positive

value is 2. The period is 6/2 = 3.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 362

362 Problems and Solutions

Problem 3. Let B1 = {|0i, |1i} and B2 = {|0H i, |1H i} denote two

orthonormal bases in C2 where
1 1
|0H i := √ (|0i + |1i), |1H i := √ (|0i − |1i).
2 2
Show that
1 1
|ψi := √ (|0i ⊗ |1i − |1i ⊗ |0i) = − √ (|0H i ⊗ |1H i − |1H i ⊗ |0H i).
2 2

Solution 3. A simple calculation yields

1
|ψi := √ (|0i ⊗ |1i − |1i ⊗ |0i)
2
1
= √ ([|0H i + |1H i] ⊗ [|0H i − |1H i] − [|0H i − |1H i] ⊗ [|0H i + |1H i])
2 2
1
= − √ (|0H i ⊗ |1H i − |1H i ⊗ |0H i).
2

Problem 4. Let

B1 := { |ψ0 i := |Hi, |ψ1 i := |V i }

denote an orthonormal basis in the Hilbert space C2 . The states |Hi and
|V i can be identified with the horizontal and vertical polarization of a
photon. Let
 
1 1
B2 := |φ0 i := √ (|Hi + |V i), |φ1 i := √ (|Hi − |V i)
2 2
denote a second orthonormal basis in C2 . These states are identified with
the 45o and -45o polarization of a photon. Alice sends photons randomly
prepared in one of the four states |Hi, |V i, |φ0 i and |φ1 i to Bob. Bob
then randomly chooses a basis B1 or B2 to measure the polarization of the
photon. All random decisions follow the uniform distribution. Alice and
Bob interpret |ψ0 i as binary 0 and |ψ1 i as binary 1 in the basis B1 . They
interpret |φ0 i as binary 0 and |φ1 i as binary 1 in the basis B2 .
(i) What is the probability that Bob measures the photon in the state
prepared by Alice, i.e. what is the probability that the binary interpretation
is identical for Alice and Bob?
(ii) An eavesdropper (named Eve) intercepts the photons sent to Bob and
then resends a photon to Bob. Eve also detects the photon polarization in
one of the bases B1 or B2 before resending. What is the probability that
the binary interpretation is identical for Alice and Bob?
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 363

Quantum Cryptography 363

Solution 4. (i) The probability that Alice chooses to prepare a state from
the basis B1 is 12 and from B2 is 12 . Similarly the probabilities that Bob
chooses to measure in the basis B1 and B2 are also 12 . Thus the probability
that Alice and Bob measure in the same basis is 14 + 14 = 12 . To determine
the correlations in the binary interpretation we consider the two cases (a)
Alice and Bob use the same basis and (b) Alice and Bob use a different
basis. The cases (a) and (b) have equal probability of 21 . For the case (a)
Alice and Bob have the same binary interpretation. For the case (b) we
note that
1
|hψ0 |φ0 i|2 = |hψ0 |φ1 i|2 = |hψ1 |φ0 i|2 = |hψ1 |φ1 i|2 = .
2
In other words, if Bob uses the wrong basis he obtains the correct binary
interpretation with probability 21 . Therefore the total probability that Alice
and Bob have the same binary interpretation is
1 1 1 3
·1+ · = .
2 2 2 4
Thus 75% of the photons sent by Alice have an identical binary interpreta-
tion shared by Alice and Bob.
(ii) From (i) the probability that Alice and Eve, Eve and Bob, as well as
Alice and Bob measure in the same basis are all 14 + 41 = 12 . Also from
(i) we find that if Alice and Eve work in the same basis Bob has a 75%
chance of obtaining the correct result since Eve does not perturb the state
of the photon. Similarly if Bob and Eve work in the same basis Bob has
a 75% chance of obtaining the correct result since Bob does not perturb
the state of the photon after Eve resends it. Now we consider the case
when Eve uses a different basis from that of Alice and Bob. Suppose Alice
sends |ψ0 i from B1 , and Eve measures in B2 . Thus Eve will obtain |φ0 i
or |φ1 i with equal probability 12 . Now Bob measures in the basis B1 and
obtains |ψ0 i with probability 21 or |ψ1 i with probability 12 . Thus we can
construct the following table where P1 is the probability that Eve obtains
Alice’s binary interpretation of the state correctly and P2 is the probability
that Bob obtains Alice’s binary interpretation of the state correctly.

Alice’s basis Eve’s basis Bob’s basis P1 P2

B1 B1 B1 1 1
B1 B1 B2 1 1/2
B1 B2 B1 1/2 1/2
B1 B2 B2 1/2 1/2
B2 B1 B1 1/2 1/2
B2 B1 B2 1/2 1/2
B2 B2 B1 1 1/2
B2 B2 B2 1 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 364

364 Problems and Solutions

The total probability that Bob’s binary interpretation corresponds to Al-

ice’s binary interpretation is
 
1 1 1 1 1 1 1 5
1+ + + + + + +1 =
8 2 2 2 2 2 2 8

i.e. 62.5%.

Problem 5. (i) Consider the two-qubit singlet state in the Hilbert space
C4
1 1
|ψi = √ (|01i − |10i) ≡ √ (|0i ⊗ |1i − |1i ⊗ |0i) .
2 2
Let U be a 2×2 unitary matrix with det(U ) = 1. Find the state (U ⊗U )|ψi.
(ii) Consider the state

1
|ψi = √ (2|0011i − |0101i − |0110i − |1001i − |1010i + 2|1100i)
2 3

in the Hilbert space C16 . This state is an extension of the two-qubit singlet
state given in (i). Calculate the state (U ⊗ U ⊗ U ⊗ U )|ψi.
(iii) The state given in (i) and (ii) can be extended to arbitrary N (N =
even) as follows
 
1 X N
|ψi = p p! − p !(−1)N/2−p |j1 j2 . . . jN i
(N/2)! N/2 + 1 permutations 2
0...01...1

where the sum is extended over all the states obtained by permuting the
state
|0 . . . 01 . . . 1i ≡ |0i ⊗ · · · ⊗ |0i ⊗ |1i ⊗ · · · ⊗ |1i
which contains the same number of 0s and 1s and p is the number of 0s in
the first N/2 positions. Thus the state is a singlet state. Let

U ⊗N ≡ U ⊗ · · · ⊗ U N − times

Find the state U ⊗N |ψi.

Solution 5. (i) A unitary transformation for 2 × 2 matrices is given by

|0i → a|0i + b|1i, |1i → c|0i + d|1i

where ad − bc = eiφ (φ ∈ R). We obtain

eiφ
(U ⊗ U )|ψi = √ (|0i ⊗ |1i − |1i ⊗ |0i) .
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 365

Quantum Cryptography 365

For φ = 0 (det(U ) = 1), we obtain the eigenvalue equation (U ⊗U )|ψi = |ψi.

(ii) Using the results from (i) and det(U ) = 1, we find

(U ⊗ U ⊗ U ⊗ U )|ψi = |ψi .

(iii) Using the result from (i), we also find U ⊗N |ψi = |ψi. The state |ψi
given in (iii) can be used to distribute cryptographic keys, encode quantum
information in decoherence-free subspaces, perform secret sharing, teleclone
quantum states, and also for solving the liar detection and Byzantine gen-
erals problems.

Problem 6. Let ρ denote an arbitrary 4 × 4 density matrix. Consider

the unitary operators (bilateral rotations)
       
1 1 i 1 i 1 1 1 1 1
Bx := ⊗ , By := ⊗ ,
2 i 1 i 1 2 −1 1 −1 1
   
1 0 1 0
Bz := i ⊗
0 −i 0 −i
and the unitary operators

U1 := I4 , U2 := Bx2 , U3 := By2 , U4 := Bz2 , U5 := Bx By , U6 := By Bz ,

U7 := Bz Bx , U8 := By Bx , U9 := U52 , U10 := U62 , U11 := U72 , U12 := U82 .

The mixed state ρW is prepared by transforming 12 systems each described
by the mixed state ρ according to each of the operators U1 , . . . , U12 . Cal-
culate the density matrix
12
1 X
ρW := Uj ρUj∗ .
12 j=1

Express ρW in terms of the Bell basis.

Solution 6. An arbitrary 4 × 4 density matrix ρ can be written in the

form
a11 a12 a13 a14
 
 a12 a22 a23 a24 
a13 a23 a33 a34
 
a14 a24 a34 1 − a11 − a22 − a33
where a11 , a12 , a13 ∈ R. We find
2 − 2F 0 0 0
 
1 0 1 + 2F 1 − 4F 0 
ρW = 
6 0 1 − 4F 1 + 2F 0

0 0 0 2 − 2F
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 366

366 Problems and Solutions

where F := (a22 + a33 − 2<a23 )/2. In terms of the Bell basis we have
1−F
ρW = F |ψ − ihψ − | + (|ψ + ihψ + | + |φ+ ihφ+ | + |φ− ihφ− |).
3

Problem 7. Let
1−F
ρW := F |φ+ ihφ+ | + (|ψ + ihψ + | + |ψ − ihψ − | + |φ− ihφ− |)
3
where F ∈ [0, 1] is the fidelity. Consider the unitary matrices
     
1 0 0 0 0 1
UBXOR1 := ⊗ I2 ⊗ I2 ⊗ I2 + ⊗ I2 ⊗ ⊗ I2 ,
0 0 0 1 1 0
     
1 0 0 0 0 1
UBXOR2 := I2 ⊗ ⊗ I2 ⊗ I2 + I2 ⊗ ⊗ I2 ⊗
0 0 0 1 1 0
and UBXOR = UBXOR1 UBXOR2 (bilateral exclusive or). Let
∗
ρ := UBXOR (ρW ⊗ ρW )UBXOR .

Calculate the probability pc that the last two qubits of a system described
by ρ are found in the same state when measured with respect to the stan-
dard basis, i.e. the last two qubits are in one of the states
       
1 1 0 0
⊗ , ⊗ .
0 0 1 1

Determine the fidelity

1 0 0 0
 
+ +
tr (ρ (|φ ihφ | ⊗ Π)) 0 0 0 0
F 0 := , Π= .
pc 0 0 0 0
0 0 0 1
It is the projection onto the space of states compatible with the measure-
ment outcomes above. Discuss the case 1/2 < F < 1.

Solution 7. The probability pc is given by

8F 2 − 4F + 5
pc = tr (ρ (I4 ⊗ Π)) = .
9
Thus the new fidelity F 0 is given by

10F 2 − 2F + 1
F0 = .
8F 2 − 4F + 5
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 367

Quantum Cryptography 367

For F = 1/2 we find F 0 = 1/2 and for F = 1 we find F 0 = 1. We have

(2F − 1)(4F − 1)(1 − F )
F0 − F = .
4F (2F − 1) + 5
Obviously for 1/2 < F < 1 we have F > 0, 2F − 1 > 0, 4F − 1 > 0 and
1 − F > 0. Thus F 0 > F when 1/2 < F < 1.

Programming Problem

Problem 1. Alice and Bob share the entangled state (one of the four
Bell states)
1
|ψi = √ (|0iA ⊗ |0iB + |1iA ⊗ |1iB ).
2
Alice applies I2 ⊗ I2 to |ψi when she wants to send the bit string 00 and
sends (I2 ⊗ I2 )|ψi. Alice applies σ3 ⊗ I2 to |ψi when she wants to send the
bit string 01 and sends (σ3 ⊗ I2 )|ψi. Alice applies σ1 ⊗ I2 to |ψi when she
wants to send the bit string 10 and sends (σ1 ⊗I2 )|ψi. Alice applies iσ2 ⊗I2
to |ψi when she wants to send the bit string 11 and sends (iσ2 ⊗ I2 )|ψi.
What states are send to Bob?

Solution 1. We have
(I2 ⊗ I2 )|ψi = ψ
1
(σ3 ⊗ I2 )|ψi = √ (|0i ⊗ |0i − |1i ⊗ |1i)
2
1
(σ1 ⊗ I2 )|ψi = √ (|1i ⊗ |0i + |0i ⊗ |1i)
2
1
(iσ2 ⊗ I2 )|ψi = √ (|0i ⊗ |1i − |1i ⊗ |0i).
2
So we have the four Bell states which form an orthonormal basis in C4 .
The corresponding Maxima program is
/* AliceBob.mac */
I2: matrix([1,0],[0,1]);
sig1: matrix([0,1],[1,0]);
sig2: matrix([0,-%i],[%i,0]);
sig3: matrix([1,0],[0,-1]);
e0: matrix([1],[0]); e1: matrix([0],[1]);
psi: (kronecker_product(e0,e0)+kronecker_product(e1,e1))/sqrt(2);
t1: kronecker_product(I2,I2) . psi;
t2: kronecker_product(sig3,I2) . psi;
t3: kronecker_product(sig1,I2) . psi;
t4: kronecker_product(%i*sig2,I2) . psi;
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 368

368 Problems and Solutions

16.3 Supplementary Problem

Problem 1. Given an orthonormal bases in C2

Bxy = { |xi, |yi }, Buv = { |ui, |vi }

with
1 1
|xi = √ (|ui + |vi), |yi = √ (|ui − |vi).
2 2
Thus we have mutually unbiased bases. Alice encodes her key-bits, for
example as a polarized photon, and sends it to Bob.
(i) Assume that Alice has chosen the state |xi with the density matrix
|xihx|. The state of Eve in C2 is |ψ0 i with the density matrix |ψ0 ihψ0 |. Eve
applies a unitary 4 × 4 matrix to the product state |xi ⊗ |ψ0 i

U (|xi ⊗ |ψ0 i) = |Bi

so that |Bi is an entangled state. Show that the Schmidt decomposition of

the state |Bi is of the form
√ √
|Bi = α|xi ⊗ |ξx i + 1 − α|yi ⊗ |ζx i

where |ξx i ⊥ |ζx i. Show that the density matrix for the post-interaction
∗
state |Xi is of the form ρAE
x = |XihX| = U (ρA E
x ⊗ ρ0 )U .
(ii) Show that when Alice sends the state |yi the entangled state

U (|yi ⊗ |ψ0 i) = |Y i

is of the form
p p
|Y i = β|yi ⊗ |ξy i + 1 − β|xi ⊗ |ζy i

where |ξy i ⊥ |ζy i.

(iii) Study the special case where U is the Bell matrix

1 0 0 1
 
1 0 1 1 0 
U=√ 
2 0 1 −1 0

1 0 0 −1

and    
1 0
|ui = , |vi = .
0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 369

Chapter 17

Quantum Channels

17.1 Introduction
We consider the Hilbert space H of n × n matrices over C with the scalar
product (Frobenius inner product)

hA, Bi := tr(AB ∗ )

with A, B ∈ H. We also consider the Hilbert space Cn and the vec oper-
ator. Given a n × m matrix the vec operator stacks the column on top of
each other.

A state is described using n × n density matrices ρ, i.e. tr(ρ) = 1 and ρ ≥ 0

(positive semidefinite). For a pure state we have ρ2 = ρ and for a mixed
state we have ρ2 6= ρ.

The space of trace-class operators acting in this Hilbert space is denoted

by S(H). A quantum channel from a Hilbert space HA to a Hilbert space
HB is represented by a completely positive trace-preserving map

Φ : S(HA ) → S(HB ).

Such a positive trace-preserving map can be represented in Stinespring

representation, Kraus operator representation and Choi-Jamiolkowski rep-
resentation.

369
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 370

370 Problems and Solutions

Let Hn denote the vector space of n × n Hermitian matrices over the real
numbers. We say that ρ ∈ Hn is positive semi-definite (or ρ ≥ 0) if x∗ ρ x ≥
0 for all x ∈ Cn , or equivalently: all of the eigenvalues of ρ are non-negative.
A linear map ψ : Hn → Hp is TPCP (trace-preserving completely positive)
if
1. TP (trace-preserving): ∀ρ ∈ Hn , tr(ρ) = tr(ψ(ρ))
2. CP (completely positive): ∀m ∈ N, ρ ∈ Hmn ,
ρ≥0 ⇒ (ψ ⊗ Im×m )(ρ) ≥ 0
where Im×m is the identity operator on m × m matrices.
Let Hn be the vector space of the n × n hermitian matrices and H ∈ Hn .
Consider a family of n × n matrices V1 , . . . , Vm over C. Consider the
completely positive map Ψ : Hn → Hn defined by
m
X
Ψ(H) = Vj HVj∗
j=1

This map is said to be a Kraus map if

m
X
Vj Vj∗ = In .
j=1

Then the matrices V1 , V2 , . . . , Vm are called Kraus operators.

A completely positive trace-preserving map Φ : S(A) → S(B) can be rep-

resented in three different ways, the Stinespring representation, Kraus op-
erator representation and Choi-Jamiolkowski representation. Stinespring’s
representation tells us that every quantum channel Φ : S(A) → S(B) can
be written in terms of an isometry V from A to the joint system B ⊗ E
(E environment) followed by a partial trace such that Φ(ρ) = trE (V ρV † )
for all ρ ∈ S(A). Tracing out system B instead of E defines a com-
plementary channel φc (ρ) = trB (V ρV † ) for all ρ ∈ S(HA ). The Choi-
Jamiolkowski representation of the channel Φ : S(A) → S(B) is the opera-
tor J(Φ) ∈ S(B ⊗ A) that is defined as
X
J(Φ) = |A|(Φ × IA )(|ΩihΩ|) = Φ(Ejk ⊗ Ejk )
1≤j,k≤|A|

where
d
1 X
|Ωi = √ |ji ⊗ |ji
d j=1
and Ejk is the elementary matrix with 1 at entry (jk) and 0 otherwise.
|A| := dim(A), |B| := dim(B) denote the input and output dimension of
the quantum channel, respectively.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 371

Quantum Channels 371

17.2 Solved Problems

Problem 1. Let Hn be the vector space of n × n hermitian matrices.
The adjoint (conjugate transpose) of a matrix A ∈ Cn×n is denoted by A∗ ,
Consider a family V1 , V2 , . . . , Vm of n × n matrices over C. We associate
with this family the completely positive map ψ : Hn → Hn defined by
m
X
ψ(X) = Vj XVj∗ .
j=1

The map ψ is said to be a Kraus map if ψ(In ) = In , i.e.

m
X
Vj Vj∗ = In
j=1

and the matrices V1 , V2 , . . . , Vm are called Kraus operators.

Let m = n = 2 and
   
0 1 0 0
V1 = , V2 = .
0 0 1 0

Show that V1 and V2 are Kraus operators and find the associated Kraus
map.

Solution 1. Since
       
0 1 0 0 0 0 0 1 1 0
V1 V1∗ + V2 V2∗ = + =
0 0 1 0 1 0 0 0 0 1

the matrices V1 and V2 are Kraus operators. The associated Kraus map is
         
a b 0 1 a b 0 0 0 0 a b 0 1
ψ = +
c d 0 0 c d 1 0 1 0 c d 0 0
 
d 0
= .
0 a

Problem 2. Let ψ : Hn → Hn be a Kraus map. Thus ψ is linear. Show

that there exists Ψ ∈ Cn×n such that for all X ∈ Hn

vec(ψ(X)) = Ψ vec(X)

where 1 is an eigenvalue of Ψ. What is a corresponding eigenvector?

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 372

372 Problems and Solutions

Solution 2. Let V1 , V2 , . . . , Vm be the Kraus operators associated to ψ.

Since vec(ABC) = (C T ⊗ A)vec(B) we find
   
X m Xm
vec(ψ(X)) = vec  Vj XVj∗  =  Vj ⊗ Vj  vec (X).
j=1 j=1

Thus we find
m
X
Ψ := V j ⊗ Vj .
j=1

We also have ψ(In ) = In so that

vec(ψ(In )) = Ψvec(In ) = vec(In )

so that 1 is an eigenvalue of Ψ and a corresponding eigenvector is vec(In ).

Problem 3. Let L denote the space of linear operators on a Hilbert space

H. A linear map  : L → L is called a positive map if for all positive semidef-
inite ρ ∈ L the operator (ρ) is also positive semidefinite. As an example we
consider the transpose operation. Let B = {|1i, |2i, . . . , |dim(H)i} denote
an orthonormal basis for the Hilbert space H. The transpose of a linear
operator
dim(H) dim(H)
X X
A := ajk |jihk|
j=1 k=1

is given by
dim(H) dim(H)
X X
T (B) (A) := ajk |kihj|.
j=1 k=1

The transpose operation T (B) is a positive map.

Let LA denote the space of linear operators on a Hilbert space HA , LB the

space of linear operators on the Hilbert space HB and LAB the space of
linear operators on the Hilbert space HA ⊗ HB . The extension  ⊗ I of the
linear map  : LA → LA is defined by
 
dim(LA ) dim(LA )
X X
( ⊗ I)  Ak ⊗ B k =
 (Ak ) ⊗ Bk
k=1 k=1

where Ak ∈ LA and Bk ∈ LB for k = 1, 2, . . . , dim(LA ). Similarly the

extension I ⊗ µ of µ : LB → LB can be defined. A positive map  : L → L
is a completely positive map if all possible extensions ( ⊗ I or I ⊗ ) of
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 373

Quantum Channels 373

the map to arbitrary Hilbert spaces are positive. Every completely positive
map can be written in the form
X
(ρ) = Ak ρA∗k
k

where Ak is a linear operator on the Hilbert space. Furthermore, if

X
Ak A∗k = I
k

then  is trace preserving, i.e. tr((ρ)) = tr(ρ). Is the transpose operation

completely positive?

Solution 3. Partial transposition T (B) ⊗ I in the basis B of the Hilbert

space HA of a state ρ in the Hilbert space HA ⊗ HB is not completely
positive. Consider the Bell state |Φ+ i in the Hilbert state C2 ⊗ C2 . Then

1 0 0 1
 
+ + 1 0 0 0 0
|Φ ihΦ | =  .
2 0 0 0 0
1 0 0 1

Thus we have
1 0 0 0
 
1 0 0 1 0
(T ({|0i,|1i}) ⊗ I)(|Φ+ ihΦ+ |) =  .

2 0 1 0 0
0 0 0 1

This last matrix has as eigenvalues 1 and −1. Consequently it is not positive
semidefinite.

Problem 4. Let |0i, |1i be an orthonormal basis in C2 . The Kraus

operators are defined by

K0 := (α|0ih0| + β|1ih1|) ⊗ I2
K1 := (β|1ih0| + α|0ih1|) ⊗ (|1ih0| + |0ih1|)

where αα∗ + ββ ∗ = 1.
(i) Show that K0∗ K0 + K1∗ K1 = I2 ⊗ I2 .
(ii) Let
1
|φi = α|0i ⊗ |0i + β|1i ⊗ |1i, |ψi = √ (|0i ⊗ |0i + |1i ⊗ |1i).
2
Show that K0 |ψihψ|K0∗ + K1 |ψihψ|K1∗ = |φihφ|.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 374

374 Problems and Solutions

Solution 4. (i) From K0∗ = (α∗ |0ih0| + β ∗ |1ih1|) ⊗ I2 we obtain

K0∗ K0 = (αα∗ |0ih0| + ββ ∗ |1ih1|) ⊗ I2 .

Since K1∗ = (β ∗ |0ih1| + α∗ |1ih0|) ⊗ (|0ih1| + |1ih0|) we have

K1∗ K1 = (ββ ∗ |0ih0| + αα∗ |1ih1)) ⊗ (|0ih0| + |1ih1|).

Since |0ih0| + |1ih1| = I2 and αα∗ + ββ ∗ = 1 we obtain the result.

(ii) Since
1
K0 |ψi = √ (α|0i ⊗ |0i + β|1i ⊗ |1i)
2
we have
A0 |ψihψ|A∗0 =
1
(αα∗ |0ih0|⊗|0ih0|+ββ ∗ |1ih1|⊗|1ih1|+αβ ∗ |0ih1|⊗|0ih1|+α∗ β|1ih0|⊗|1ih1|).
2
Since
1
A1 |ψi = √ (α|0i ⊗ |0i + β|1i ⊗ |1i) = A0 |ψi
2
∗
we obtain A1 |ψihψ|A1 = A0 |ψihψ|A0 . Thus

A0 |ψihψ|A∗0 + A1 |ψihψ|A∗1 =

αα∗ |0ih0| ⊗ |0ih0| + ββ ∗ |1ih1| ⊗ |1ih1| + αβ ∗ |0ih1| ⊗ |0ih1| + α∗ β|1ih0| ⊗ |1ih0|

which is the density matrix |φihφ|.

Problem 5. Find all Kraus maps ψ : H2 → H2 , associated with families

of 2 Kraus operators (V1 and V2 ), which provide the transformation
   
1 0 0 0
ψ = .
0 0 0 1
Calculate  
0 0
ψ .
0 1
Is there a Kraus map associated with a single Kraus operator which also
provides this transformation?

Solution 5. By linearity of the Kraus map ψ we have

           
1 0 1 0 1 0 0 0 0 0 0 0
=ψ =ψ +ψ = +ψ .
0 1 0 1 0 0 0 1 0 1 0 1
Thus    
0 0 1 0
ψ = .
0 1 0 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 375

Quantum Channels 375

We construct the matrix Ψ which implements ψ under the vec operator.

We write Ψ = Π1 + A(I4 − Π1 ), where A is a 4 × 4 matrix such that ψ is a
Kraus map providing the given transformation and
1 0 0 1
 
1 1 0 0 0 0
Π1 := vec(I2 )(vec(I2 ))∗ = 
2 2 0 0 0 0

1 0 0 1
is a projection matrix. We have
1 1 1 0
       
   
1 0 0 1 0 1  0  0 0 0
Ψvec = Ψ  =  + A = = vec
0 0 0 2 0 2 0 0 0 1
  
0 1 −1 1
so that
1 −1
   
 0   0 
A = .
0 0
−1 1
We write A = −Π−1 + B(I4 − Π−1 ), where B is a 4 × 4 matrix and
∗
1 1 1 0 0 −1
   
1  0  0  1 0 0 0 0 
Π−1 :=   = 
2 0 0 2 0 0 0 0
 
−1 −1 −1 0 0 1
is a projection operator with Π−1 Π1 = Π1 Π−1 = 04 . Consequently,
0 0 0 1 0 0 0 0 0 b12 b13 1
     
0 0 0 0  0 1 0 0   0 b22 b23 0 
Ψ= +B = .
0 0 0 0 0 0 1 0 0 b32 b33 0
1 0 0 0 0 0 0 0 1 b42 b43 0
Clearly Ψ cannot be written as a Kronecker product of 2×2 matrices, so no
Kraus map associated with a single Kraus operator can provide the given
map. Assume Ψ = C ⊗ C + D ⊗ D for some 2 × 2 matrices C and D. We
find the equations
   
0 b12 b13 1
= c11 C + d11 D, = c12 C + d12 D,
0 b22 b23 0
   
0 b32 b33 0
= c21 C + d21 D, = c22 C + d22 D.
1 b42 b43 0
Clearly the first columns (second columns) of C and D are linearly depen-
dent. Thus we find
   
0 c2 0 d2
V1 = C = , V2 = D =
c1 0 d1 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 376

376 Problems and Solutions

where c1 , c2 , d1 , d2 ∈ C satisfy |c1 |2 + |d1 |2 = 1, |c2 |2 + |d2 |2 = 1.

Problem 6. Let ρ1 , ρ2 ∈ Hn be positive semi-definite matrices.

(i) Is ρ3 = ρ1 + ρ2 positive semi-definite?
(ii) Is ρ4 = kρ1 (k ∈ C) positive semi-definite?
(iii) Let ρ5 ∈ Hn such that ρ1 + ρ5 is positive semi-definite. Is ρ5 positive
semi-definite?

Solution 6. (i) Let x ∈ Cn , then x∗ ρ3 x = x∗ ρ1 x + x∗ ρ1 x ≥ 0 since

x∗ ρ1 x ≥ 0 and x∗ ρ1 x ≥ 0. Thus ρ3 is positive semi-definite.
(ii) Let x ∈ Cn , then x∗ ρ4 x = kx∗ ρ1 x is non-negative if and only if k ∈ R
with k ≥ 0. Thus, in general, ρ4 = kρ1 is positive semi-definite if and only
if k ≥ 0 (the only exception is when ρ1 is the zero matrix, in which case
∈ C).  
the statement is true for all k    
1 1 1 0 0 1
(iii) In general, no. Consider = + . All of the
1 1 0 1 1 0
matrices
  appearing in this equation are positive semi-definite, except for
0 1
.
1 0

Problem 7. Show that a linear map ψ : Hn → Hp is a TP map if

∗
and only if ψ ∗ (In ) = In , where denotes the adjoint with respect to the
Frobenius inner product and In is the n × n identity matrix.

Solution 7. The Frobenius inner product hA, Bi := tr(B ∗ A) provides

tr(ρ) = tr(In∗ ρ) = hρ, In i
tr(ψ(ρ)) = tr(I ∗ ψ(ρ)) = hψ(ρ), In i = hρ, ψ ∗ (In )i
so that
tr(ρ) = tr(ψ(ρ)) ⇔ ψ ∗ (In ) = In .

Problem 8. Show that

n
X
ρ0 := (Eij ⊗ Eij ) ∈ Hn2
i,j=1

is positive semi-definite, where Eij is the elementary n × n matrix with a 1

in row i and column j and 0 elsewhere.

2
Solution 8. Any vector x in Cn can be written in the form
n
X
x= xs,t es ⊗ et
s,t=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 377

Quantum Channels 377

where {e1 , . . . , en } is the standard basis in Cn and xs,t ∈ C for s, t ∈

{1, . . . , n}. Since e∗s Eij = δis e∗j we find

X 2
n
 n 
X
∗
x ρ0 x = xii xjj = xii  ≥ 0.
 
 
i,j=1 i=1

Problem 9. An orthonormal basis, with respect to the Frobenius inner

product, for Hn (n ≥ 2) is given by B = B1 ∪ B2 where
 
1
B1 = √ (Ejk + Ekj ) : j, k = 1, . . . , n, j ≤ k
2
 
i
B2 = √ (Ejk − Ekj ) : j, k = 1, . . . , n, j < k .
2

Express
n
X
ρ0 := (Eij ⊗ Eij ) ∈ Hn2
i,j=1

in terms of this basis.

Solution 9. We find
X X
ρ0 = F ⊗F − G ⊗ G.
F ∈B1 G∈B2

Problem 10. Show that a linear map ψ : Hn → Hp is a CP map if and

only if (ψ ⊗ In×n )(ρ0 ) is positive semi-definite where
n
X
ρ0 := (Eij ⊗ Eij ) ∈ Hn2 .
i,j=1

Solution 10. If ψ is completely positive, it follows immediately that

(ψ ⊗ In×n )(ρ0 ) is positive semi-definite (since ρ0 is positive semi-definite).
Consider the linear extension ψe : Mn×n → Mp×p of ψ, i.e. ψ(ρ) e = ψ(ρ) for
all ρ ∈ Hn . Assume that
n
X
(ψ ⊗ In×n )(ρ0 ) = e ij ) ⊗ Eij
ψ(E
i,j=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 378

378 Problems and Solutions

is positive semi-definite. This is the Choi-Jamiolkowski representation. We

have a spectral decomposition
2 2
n
X n
X
(ψ ⊗ In×n )(ρ0 ) = λk uj u∗k = vk vk∗
k=1 k=1
√
where {u1 , . . . , un } is an orthonormal set and vk := λk uk (note that
λk ≥ 0). Define
Pj := In ⊗ e∗j
and consider Pi (ψ ⊗ In×n )(ρ0 )Pj∗ :
2
n
X n
X
(ψ ⊗ In×n )(ρ0 ) = e ij ) ⊗ Eij =
ψ(E vk vk∗
i,j=1 k=1

2
n
X
Pi (ψ ⊗ In×n )(ρ0 )Pj∗ = ψ(E
e ij ) = Pi vk (Pj vk )∗
k=1

and define Vk ej := Pj vk so that

2
n
X
Pi (ψ ⊗ In×n )(ρ0 )Pj∗ = ψ(E
e ij ) = Vk Eij Vk∗ .
k=1

By linear extension, the map ψ is given by

2
n
X
ψ(ρ) := Vk ρVk∗
k=1

which is completely positive since for ρ ∈ Hmn , ρ ≥ 0

2
n
X
(ψ ⊗ Im×m )ρ = (Vk ⊗ Im )ρ(Vk ⊗ Im )∗
k=1

is positive semi-definite. Consequently ψ is completely positive if and only

if there exists V1 , . . . Vn2 (Kraus operators) such that
2
n
X
ψ(ρ) := Vk ρVk∗ .
k=1

This is a Kraus representation.

Problem 11. Is the map ψ : Hn → Hp given by ψ(ρ) = ρT completely

positive?
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 379

Quantum Channels 379

Solution 11. Consider

n
X
ρ0 := (Eij ⊗ Eij ) ∈ Hn2 .
i,j=1

We have
n
X n
X
(ψ ⊗ In×n )(ρ0 ) = e ij ) ⊗ Eij =
ψ(E Eji ⊗ Eij .
i,j=1 i,j=1
2
Any vector x in C n can be written in the form
n
X
x= xs,t es ⊗ et
s,t=1

where {e1 , . . . , en } is the standard basis in Cn and xs,t ∈ C for s, t ∈

{1, . . . , n}. Since e∗s Eij = δis e∗j we find
n
X
x∗ (ψ ⊗ In×n )(ρ0 )x = xji xij 6≥ 0
i,j=1

in general (consider x12 = −x21 = 1 and all other coefficients are zero).

Problem 12. Let ψ : Hn → Hp given by

2
n
X
ψ(ρ) := Vk ρVk∗
k=1

be a CP map. Find the condition on V1 , . . . , Vn2 such that ψ is TP (and

hence TPCP).

Solution 12. The adjoint of

2
n
X
ψ(ρ) := Vk ρVk∗
k=1

with respect to the Frobenius inner product is

2
n
X
ψ ∗ (ρ) := Vk∗ ρVk
k=1

and since ψ is TP if and only if ψ ∗ (In ) = In we find

2
n
X
tr(ρ) = tr(ψ(ρ)) ⇔ Vk∗ Vk = In .
k=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 380

380 Problems and Solutions

Problem 13. Let ψ : Hn → Hp given by

2
n
X
ψ(ρ) := Vk ρVk∗
k=1

be a CP map. Show that there exists a matrix V such that

ψ(ρ) = trm (V (ρ ⊗ In2 )V ∗ ).

Solution 13. Let

2
n
X
V = (Vk ⊗ Ekk ).
k=1

Then
2 2
n
X n
X
∗
V (ρ ⊗ Im )V = δkl (Vk ρVl∗ ) ⊗ Ekk = (Vk ρVk∗ ) ⊗ Ekk
k,l=1 k=1

and
2
n
X
∗
trn2 (V (ρ ⊗ In2 )V ) = (Vk ρVk∗ )tr(Ekk ) = ψ(ρ).
k=1

This is a Stinespring representation.

Problem 14. A minimal Stinespring representation of a CP map ψ :

Hn → Hp is a representation

ψ(ρ) = trm (V (ρ ⊗ Im )V ∗ )

where m is minimal. This corresponds to minimizing the number of non-

zero Kraus operators Vk in a Kraus representation
m
X
ψ(ρ) := Vk ρVk∗ .
k=1

Given a Kraus representation

2
n
X
ψ(ρ) := Vek ρVek∗ .
k=1

Consider
2
n
X
A= (vec(Vek ))(vec(Vek ))∗ .
k=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 381

Quantum Channels 381

The matrix A is positive definite and thus has a spectral decomposition

m
X
A= λk vk vk ∗
k=1

where m is the rank of A and λ1 , . . . , λm > 0. We √

find the Kraus operators
Vk for the minimal representation from vecVk = λk vk . Find a minimal
representation for the completely map on H2 given by
       
0 1 0 1 0 0 0 0
V1 =
e , V2 =
e , V3 =
e , V4 =
e .
1 0 0 0 1 0 0 0

Solution 14.The matrix A is given by

0 0 0 0
 
0 2 1 0
A=
0 1 2 0

0 0 0 0
0 0
   
1 1 1 1  1  1
=3 · √   √ (0 1 1 0) + 1 · √   √ (0 1 −1 0)
2 1 2 2 −1 2
0 0
with non-zero eigenvalues 1 and 3. An optimal Kraus representation is
given by r    
3 0 1 1 0 −1
V1 = , V2 = √ .
2 1 0 2 1 0
We verify that the map is correct
   
a b 2d c
ψ =
c d b 2a
   
3 0 1 a b 0 1
=
2 1 0 c d 1 0
   
1 0 −1 a b 0 1
+ .
2 1 0 c d −1 0
The corresponding minimal Stinespring representation is given by
n r        
X 3 0 1 1 0 1 0 −1 0 0
V = Vk ⊗ Ekk = ⊗ +√ ⊗
2 1 0 0 0 2 1 0 0 1
k=1
√
0 0 3 0
 
1  0 0 0 −1 
= √ √
3 0 0 0

2
0 1 0 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 382

382 Problems and Solutions

i.e.
     
a b a b 1 0
ψ = tr2 V ⊗ V∗
c d c d 0 1
3d 0 3c 0
 
 
1 0 d 0 −c  2d c
= tr2  = .
2 3b 0 3a 0 b 2a

0 −b 0 a

Problem 15. Consider the Kraus operators K1 and K2

       
0 1 0 0 0 0 0 1
K1 = ⇒ K1∗ = , K2 = ⇒ K2∗ =
0 0 1 0 1 0 0 0

and an arbitrary 2 × 2 matrix A = (ajk ). Then

 
a22 0
K1 AK1∗ + K2 AK2∗ = .
0 a11

So the trace is preserved under this transformation. Let

c†1 , c†2 , c1 , c2

be Fermi creation and annihilation operators, respectively. Consider the

operators  

0 1
K̂1 = c†1 c†2 = c†1 c2 , K̂1† = c†2 c1
0 0
 

0 0
K̂2 = c†1 c†2 = c†2 c1 , K̂2† = c†1 c2
1 0
and
 
c1
c†1 c†2 = a11 c†1 + a12 c†1 c2 + a21 c†2 c†2 c1 + a22 c†2 c2 .


 = A
c2

Find the operator K̂1 ÂK̂1† + K̂2 ÂK̂2† .

Solution 15. With c†j ck + ck c†j = δjk I we obtain

K̂1 ÂK̂1† + K̂2 ÂK̂2† = a11 c†2 c2 + a22 c†1 c1 − (a11 + a22 )c†1 c1 c†2 c2 )
  

a22 0 c1
= c†1 c†2 − (a11 + a22 )c†1 c1 c†2 c2 .
0 a11 c2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 383

Quantum Channels 383

Programming Problem

Problem 1. Let µ ∈ [0, 1]. Apply computer algebra to show that the
2 × 2 matrices

√ √ √ √
1 + 3µ 1−µ 1−µ 1−µ
K0 = I2 , K1 = σ1 , K2 = σ2 , K3 = σ3
2 2 2 2

are Kraus operators. Show that the sixteen 4 × 4 matrices Kj ⊗ K` , (j, ` =

0, 1, 2, 3) are Kraus operators.

Solution 1. Note that the operators Kj and Kj ⊗ K` are hermitian. The

Maxima program is

/* Kraus.mac */
I2: matrix([1,0],[0,1]);
sig1: matrix([0,1],[1,0]);
sig2: matrix([0,-%i],[%i,0]);
sig3: matrix([1,0],[0,-1]);
K0: sqrt(1+3*mu)*I2/2; K1: sqrt(1-mu)*sig1/2;
K2: sqrt(1-mu)*sig2/2; K3: sqrt(1-mu)*sig3/2;
K0TC: K0; K1TC: K1; K2TC: K2; K3TC: K3;
S: K0 . K0TC + K1 . K1TC + K2 . K2TC + K3 . K3TC;
S: ratsimp(S);
K00: kronecker_product(K0,K0); K01: kronecker_product(K0,K1);
K02: kronecker_product(K0,K2); K03: kronecker_product(K0,K3);
K10: kronecker_product(K1,K0); K11: kronecker_product(K1,K1);
K12: kronecker_product(K1,K2); K13: kronecker_product(K1,K3);
K20: kronecker_product(K2,K0); K21: kronecker_product(K2,K1);
K22: kronecker_product(K2,K2); K23: kronecker_product(K2,K3);
K30: kronecker_product(K3,K0); K31: kronecker_product(K3,K1);
K32: kronecker_product(K3,K2); K33: kronecker_product(K3,K3);
K00TC: K00; K01TC: K01; K02TC: K02; K03TC: K03;
K10TC: K10; K11TC: K11; K12TC: K12; K13TC: K13;
K20TC: K20; K21TC: K21; K22TC: K22; K23TC: K23;
K30TC: K30; K31TC: K31; K32TC: K32; K33TC: K33;
S: K00 . K00TC + K01 . K01TC + K02 . K02TC + K03 . K03TC
+ K10 . K10TC + K11 . K11TC + K12 . K12TC + K13 . K13TC
+ K20 . K20TC + K21 . K21TC + K22 . K22TC + K23 . K23TC
+ K30 . K30TC + K31 . K31TC + K32 . K32TC + K33 . K33TC;
S: ratsimp(S);
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 384

384 Problems and Solutions

17.3 Supplementary Problems

3
Problem
√ √ |0i, |1i, |2i be the standard basis in C and α = (2 −
1. Let
3)/4, β = 2 + 3. Show that
√ p p
K1 = α( β|0ih0| + |1ih1|) ⊗ ( β|0ih1| + |1ih0|)
√ p p
K2 = α(|1ih0| + β|0ih1|) ⊗ (|1ih1| + β|0ih0|)
√ p p
K3 = α( β|0ih0| + |1ih2|) ⊗ ( β|0ih2| + |1ih0|)
√ p p
K4 = α(|1ih0| + β|0ih2|) ⊗ (|1ih2| + β|0ih0|)
p
K5 = |1ih1| ⊗ (2 αβ|1ih1| + |2ih2|)
p
K6 = |2ih2| ⊗ (|1ih1| + 2αβ|2ih2|)
are Kraus operators in C9 .

Problem 2. Let µ ∈ [0, 1] and σ1 , σ2 , σ3 , σ0 = I2 be the Pauli spin

matrices. Then the four 2 × 2 matrices
√ √ √ √
1 + 3µ 1−µ 1−µ 1−µ
K0 = σ0 , K1 = σ1 , K2 = σ2 , K3 = σ3
2 2 2 2
are Kraus operators. Show that the sixteen 4 × 4 matrices Kj ? K` , (j, ` =
0, 1, 2, 3) are Kraus operators, where ? denotes the star product.

Problem 3. Let Kj (j = 1, . . . , m) be n × n matrices over C with

m
X
Kj Kj∗ = In .
j=1

Show that
m X
X m
(Kj ⊗ K` )(Kj∗ ⊗ K`∗ ) = In ⊗ In ≡ In2 .
j=1 `=1

Problem 4. Consider the 4 × 4 matrices with trace 1

1 0 0 1 0 0 0 0
   
1 0 0 0 0 1 0 1 1 0
A=  , B =  .
2 0 0 0 0 2 0 1 1 0
1 0 0 1 0 0 0 0
(i) Can one find a unitary matrix U such that U AU −1 = B?
(ii) Apply the vec operator to A and B. Then find a 16 × 16 unitary matrix
W such that W vec(A) = vec(B).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 385

Part II

Infinite-Dimensional
Hilbert Spaces

385
August 1, 2014 10:17 9in x 6in Scanning Transmission Electron Microscopy. . . b1683-fm

This page intentionally left blank

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 387

Chapter 18

Bose Operators and

Number States

18.1 Introduction
Besides qubit-based quantum computing and quantum algorithms, quan-
tum information over continuous variables is also applied and used in fields
such as quantum teleportation and quantum cryptography. For continuous
systems Bose operators play the central role. Consider a family of linear
operators bj , b†j , j = 1, 2, . . . , m on an inner product space V , satisfying the
commutation relations (Heisenberg algebra)

[bj , bk ] = [b†j , b†k ] = 0, [bj , b†k ] = δjk I (1)

where I is the identity operator. The operator b†j is called a Bose creation
operator and the operator bj is called an Bose annihilation operator. The
inner product space must be infinite dimensional for (1) to hold. For, if A
and B are n × n matrices such that [A, B] = λI, then

tr([A, B]) = 0

implies λ = 0.

Let |0i ≡ |00...0i be the vacuum state, i.e.

bj |0i = 0|0i

387
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 388

388 Problems and Solutions

with h0|0i = 1, j = 1, 2, . . . , m. A normalized state is given by

1
√ (b† )n1 (b†j2 )n2 · · · (b†jk )nk |0i ≡ |n1 , n2 , . . . , nk i
n1 !n2 ! · · · nk ! j1
where j1 , j2 , . . . , jk ∈ { 1, 2, . . . , m } and n1 , n2 , . . . , nk ∈ {0, 1, 2, . . . }. The
states are called number states (also called Fock states). For bj we also use
the notation I ⊗ I ⊗ · · · ⊗ b ⊗ I ⊗ I ⊗ · · · ⊗ I, where b is in the jth position.

Let m = 1. Consider the number states |ni (n = 0, 1, . . .). Then

√ √
b|ni = n|n − 1i, b† |ni = n + 1|n + 1i, b† b|ni = n|ni
and
∞
X
|nihn| = I
n=0

where I is the identity operator and n̂ = b† b is the number operator. We

have
1
|ni = √ (b† )n |0i, n = 0, 1, 2, . . . .
n!
The number states |ni form an orthonormal basis in the Hilbert space
`2 (N0 ). The Hamilton operator Ĥ of the one-dimensional harmonic oscil-
lator can be written as
 
† 1
Ĥ = ~ω b b + I
2

with the eigenvalue equation Ĥ|ni = (~ω + 1/2)|ni

Consider the commutation relation

[q̂, p̂] = i~I
with p̂ = −i~∂/∂q. The electric field E for a single mode in a box of length
L can be written as
r
~ω  −i(ωt−k·x) † i(ωt−k·x)

E(x, t) = ieλ b(0)e − b (0)e
20 L3
where k · x = k1 x1 + k2 x2 + k3 x3 .

Define the unitary operators

Ûα = exp(iαp̂/~), V̂α = exp(iβ q̂/~).
Then
Ûα V̂β = exp(iαβ/~)V̂β Ûα
with α dimension meter and β dimension kg . meter . sec−1 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 389

Bose Operators and Number States 389

18.2 Solved Problems

Problem 1. Consider the Hamilton operator for the one dimensional
harmonic oscillator
1 2 1
Ĥ = p̂ + mω 2 q̂ 2
2m 2
where ω is the frequency and m the mass. We introduce the characteristic
length r
~
`0 :=
mω
and define the dimensionless linear unbounded operators (Bose operators)
   
1 q̂ p̂ 1 q̂ p̂
b := √ +i , b† := √ −i .
2 `0 ~/`0 2 `0 ~/`0

(i) Find the commutator [b, b† ].

(ii) Express q̂ and p̂ in terms of b and b† , i.e. find the inverse transform.
(iii) Express Ĥ in terms of b and b† . Find Ĥ|ni. Discuss

Solution 1. (i) Since

∂
p̂ := −i~
∂q
we obtain [b, b† ] = I, where I is the identity operator.
(ii) We find
1 ~/`0
q̂ = √ `0 (b + b† ), p̂ = √ (b − b† ).
2 2i
(iii) We obtain
1
Ĥ = ~ω(b† b + I)
2
and the eigenvalue equation Ĥ|ni = ~ω(n + 1/2)|ni with n = 0, 1, . . ..

Problem 2. Let b, b† be Bose annihilation and creation operators. Find

the commutator of the operators
eb = 1 I + b, eb† = 2 I + b†

where I is the identity operator and 1 , 2 are constants.

Solution 2. We have

[eb, eb† ] = [1 I + b, 2 I + b† ] = [1 I, 2 I] + [1 I, b† ] + [b, 2 I] + [b, b† ]

= [b, b† ]
= I.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 390

390 Problems and Solutions

Problem 3. (i) Let β ∈ C. Calculate the commutators

[b, βb† − β ∗ b], [b† , βb† − β ∗ b].

(ii) Let n ∈ N and β ∈ C. Calculate the commutator [b† , (βb† − β ∗ b)n ].

Solution 3. (i) We find

[b, βb† − β ∗ b] = βI, [b† , βb† − β ∗ b] = β ∗ I.

(ii) If f is an analytic function we have

df (b)
[b† , f (b)] = − .
db
Thus
[b† , (βb† − β ∗ b)n ] = nβ ∗ (βb† − β ∗ b)n−1 .
For n = 1 we have [b† , (βb† − β ∗ n)] = β ∗ I.

Problem 4. Let n̂ := b† b and α ∈ R. Calculate the state exp(αn̂)|0i.

Solution 4. Since n̂|0i = 0|0i we obtain exp(αn̂)|0i = |0i.

Problem 5. (i) Calculate the commutators [b2 , b† b], [b2 , b†2 ].

(ii) Using the commutation relations and b|0i = 0|0i calculate the state

bbb† b† |0i.

Is the state normalized?

(iii) Let
1
|ni := √ (b† )n |0i, n = 0, 1, 2, . . .
n!
be the number states. Find the operator
∞
X
|nihn|.
n=0

Solution 5. (i) Using the commutation relations we obtain

[b2 , b† b] = 2b2 , [b2 , b†2 ] = 2I + 4b† b.

(ii) Using the commutation relations and b|0i = 0|0i we find

bbb† b† |0i = 2|0i.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 391

Bose Operators and Number States 391

(iii) Since |ni (n = 0, 1, 2, . . .) is an orthonormal basis we find

∞
X
|nihn| = I
n=0

where I is the identity operator, i.e. we have the infinite-dimensional unit

matrix. This is the completeness relation.

Problem 6. Using the number states |ni find the matrix representation
of the unbounded operators b† b.

Solution 6. Since b† b|ni = n|ni we obtain the infinite dimensional un-

bounded diagonal matrix diag(0, 1, 2, . . .).

Problem 7. Let n̂ := b† b be the number operator. Calculate the commu-

tators [n̂, b], [n̂, [n̂, b]], [n̂, [n̂, [n̂, b]]]. Discuss the general case for m commu-
tators.

Solution 7. We have [n̂, b] = −b, [n̂, [n̂, b]] = (−1)2 b, [n̂, [n̂, [n̂, b]]] =
(−1)3 b. Obviously, for the general case with m commutators we find
(−1)m b.

Problem 8. Using the number states |ni find the matrix representation
of the unbounded operators b† + b.

√ √
Solution 8. Since b|ni = n|n − 1i, b† |ni = n + 1|n + 1i we obtain the
infinite dimensional unbounded symmetric matrix
0 1 √0 0 ...
 
 1 √0 2 √0 . . . 
0 2 √0 3 ....
 
0 0 3 0 ...
 
.. .. .. ..
. . . .

Problem 9. Let b† , b be Bose creation and annihilation operators, re-

spectively.
(i) Calculate the commutator [b† + b, b† b].
(ii) Consider the symmetric 4 × 4 matrices
0 0 0 0 0 1 √0 0
   
0 1 0 0  1 √0 2 √0 
A= , B= .
0 0 2 0 0 2 √0 3
0 0 0 3 0 0 3 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 392

392 Problems and Solutions

Find the commutator [A, B]. These matrices appear when we truncate the
infinite dimensional unbounded matrices b† b and b† + b.

Solution 9. (i) Using bb† = I + b† b we obtain [b† + b, b† b] = −b† + b.

(ii) We obtain the skew-symmetric matrix

0 −1 0 0
 
√
1 √0 − 2 0 
√
[A, B] = 
0 2 √0 − 3

0 0 3 0

which is a truncation of the infinite dimensional unbounded matrix −b† + b.

Problem 10. (i) Let  ∈ R. Find

† † † †
fc () = e−ib b b† eib b , fa () = e−ib b beib b

i.e. fc (0) = b† , fa (0) = b.

(ii) Then find the 2 × 2 matrix A() such that
† †
e−ib b (b† + b)eib b b† + b
   
= A() .
ib† − ib
† †
e−ib b (ib† − ib)eib b

Solution 10. Applying parameter differentiation and bb† = I + b† b we

obtain the differential equation

dfc
= −ifc
d

with the initial condition fc (0) = b† . The solution is fc () = e−i b† . Anal-
ogously for fa we find dfa /d = ifa with the initial condition fa (0) = b.
Hence fa () = ei b.
(ii) Since
† †
e−ib b (b† + b)eib b = e−i b† + ei b
† †
e−ib b (ib† − ib)eib b = ie−i b† − iei b

we obtain
 −i †
(cos() − i sin())b† + (cos() + i sin())b
  
e b + ei b
=
ie−i b† − iei b i(cos() − i sin())b† − i(cos() + i sin())b
  † 
cos() − sin() b +b
= .
sin() cos() ib† − ib
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 393

Bose Operators and Number States 393

Problem 11. Consider the operators

 = bk e−iφk + b†k eiφk , B̂ = b` e−iφ` + b†` eiφ` .

Find the commutator [Â, B̂]. What is the condition on the phases φk and
φ` such that [Â, B̂] = 0 for k = `?

Solution 11. We have

[Â, B̂] = [bk e−iφk + b†k eiφk , b` e−iφ` + b†` eiφ` ]

= [bk e−iφk , b†` eiφ` ] + [b†k eiφk , b` e−iφ` ]
= (ei(φ` −φk ) − ei(φk −φ` ) )δ`k
= 2i sin(φ` − φk )δ`k .

Thus the commutator is zero for k = ` if sin(φ` − φk ) = 0 i.e. φ` − φk = nπ

and n ∈ Z.

Problem 12. Let f be an analytic function in z and z ∈ C. Let n̂ := b† b

be the number operator. Calculate
†
ezn̂ f (b), ezn̂ f (b† ), ezb f (b), ezb f (b† ).

Solution 12. Since n̂b = b(n̂ − I) and n̂b† = b† (I + n̂) we find

ezn̂ f (b) = f (e−z b)ezn̂

ezn̂ f (b† ) = f (ez b† )ezn̂
† †
ezb f (b) = f (b − zI)ezb
ezb f (b† ) = f (b† + zI)ezb .

Problem 13. Let b† , b be Bose creation and annihilation operator with

the commutator relation [b, b† ] = I, where I is the identity operator. Show
that one has a representation

d
b† 7→ z, b 7→ .
dz

Solution 13. We calculate the commutator

[d/dz, z]f (z)

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 394

394 Problems and Solutions

where f is an analytic functions. Using the product rule we obtain

d df (z) df (z) df (z)

[d/dz, z]f (z) = (zf (z)) − z = f (z) + z −z = f (z).
dz dz dz dz

Problem 14. Consider the differential operators

   
1 d 1 d
b= √ x+ , b† = √ x−
2 dx 2 dx

acting in the vector space S(R). Find the commutator [b, b† ]. Find the
operator N̂ = b† b.

Solution 14. Let f ∈ S(R). We have

[b, b† ]f = bb† f − b† bf
d2 f
 
1 2 d df
= x f− (xf ) − x −
2 dx dx dx2
d2 f
 
1 2 d df
− x f− (xf ) + x −
2 dx dx dx2
d df
= (xf ) − x
dx dx
= f.

We obtain the differential operator

d2
 
1
N̂ = x2 − 1 − .
2 dx2

Problem 15. Let  ∈ R and  > 0, n̂ = b† b and { |ni : n = 0, 1, 2, . . . }

denote the number states.
(i) Using the number states calculate the trace
∞
X
tr(e−n̂ ) ≡ hn|e−n̂ |ni.
n=0

(ii) Using the number states calculate the trace

∞
† X †
tr(b† be−b b ) = hn|b† be−b b |ni.
n=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 395

Bose Operators and Number States 395

†
Solution 15. (i) Since n̂|ni = n|ni we have e−b b |ni = e−n |ni. Thus
∞
X e
tr(e−n̂ ) = e−n = .
n=0
e −1

(ii) Using hn|b† b = hn|n we obtain

∞
† X e
tr(b† be−b b ) = ne−n = .
n=0
(e − 1)2

Problem 16. Consider the Hamilton operator Ĥ = ~ωb† b.

(i) Calculate the trace tr(e−Ĥ/kB T ).
(ii) Consider the density operator

exp(−Ĥ/kB T )
ρ= .
tr(exp(−Ĥ/kB T ))

Let n̂ := b† b. Calculate n ≡ hn̂i ≡ hb† bi ≡ tr(b† bρ) and hĤi.

(iii) Calculate tr(ρb), tr(ρb† ).

Solution 16. (i) We set λ := ~ω/(kB T ). Then using the completeness

relation
∞
X
I= |nihn|,
n=0

n̂|ni = n|ni and e−λn̂ |ni = e−λn |ni we find

∞ ∞
† X † X
e−Ĥ/kB T = e−λb b
= e−λb b |nihn| = e−λn |nihn|.
n=0 n=0

Thus
∞
X 1
tr(e−Ĥ/kB T ) = (e−λ )n = .
n=0
1 − e−λ
(ii) Using the result from (i) we obtain
1
n = hb† bi =
eλ −1
and therefore
~ω
hĤi = .
e~ω/kB T − 1
(iii) Since hn|b|ni = 0, hn|b† |ni = 0 we obtain

hbi = tr(ρb) = 0, hb† i = tr(ρb† ) = 0.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 396

396 Problems and Solutions

Problem 17. Consider the Bose-Einstein density operator

∞  n
1 X n
ρ= |nihn|
n + 1 n=0 n + 1

where |ni, n = 0, 1, 2, . . . are the number states and

   −1
~ω
n := exp −1 .
kB T

Show that tr(ρ) = 1.

Solution 17. We set α = ~ω/(kB T ) and use the number states as an

orthonormal basis. Thus
∞ ∞  n
X 1 X n
tr(ρ) = hm|ρ|mi =
m=0
n + 1 n=0 n + 1

where we used hm|ni = hn|mi = δmn . Now we have

1 n
= 1 − e−α , = e−α .
n+1 n+1
Since the sum is a geometric series we have
∞
X 1
(e−α )n = .
n=0
1 − e−α

Thus tr(ρ) = 1.

Problem 18. Let b and b† be Bose annihilation and creation operators,

respectively. Consider the general one-mode canonical Bogolubov transform

eb := eiφ cosh(s)b + eiψ sinh(s)b†

eb† := e−iφ cosh(s)b† + e−iψ sinh(s)b

where s is a real parameter (squeezing parameter).

(i) Show that the operators eb and eb† satisfy the Bose commutation relations.
(ii) Find the inverse Bogolubov transform.

Solution 18. (i) Since cosh2 (s) − sinh2 (s) = 1 we find

[eb, eb† ] = bb† − b† b = I.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 397

Bose Operators and Number States 397

(ii) The transformation can be written in the matrix form

    
eb eiφ cosh(s) eiψ sinh(s) b
= −iψ .
eb† e sinh(s) e−iφ cosh(s) b†

The determinant of this matrix is +1. Thus the inverse transformation is

given by
   −iφ  
b e cosh(s) −eiψ sinh(s) eb
= eb† .
b† −e−iψ sinh(s) eiφ cosh(s)

Problem 19. Let  ∈ R and f : C2 → C be an entire analytic function.

If a function f is analytic on the whole complex plane then f is said to be
entire. Show that
eb f (b, b† )e−b = f (b, b† + I) (1a)
† †
e−b f (b, b† )eb = f (b + I, b† ). (1b)

Solution 19. We have

eb f (b, b† )e−b = f (eb be−b , eb b† e−b ) = f (b, eb b† e−b ).

Since eb b† e−b = b† + I we find (1a), where we used [b, b† ] = I. A similar

proof holds for (1b).

Problem 20. Let f : C → C be an entire analytic function. Show that

∞
X
f (b† b) = f (n)|nihn|
n=0

where |ni is the number state.

Solution 20. The completeness relation is given by

∞
X
|nihn| = I.
n=0

Since b† b|ni = n|ni, we have f (b† b)|ni = f (n)|ni. It follows that

∞
X ∞
X ∞
X
f (b† b) = f (b† b)I = f (b† b) |nihn| = f (b† b)|nihn| = f (n)|nihn|.
n=0 n=0 n=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 398

398 Problems and Solutions

Problem 21. Let f be an analytic function in x and y. Let b† and b

be Bose creation and annihilation operators, respectively. We can define
f (b, b† ) by its power series expansion

X ∞
∞ X ∞
X
f (b, b† ) := ··· f (j1 , j2 , . . . , jn )(b† )j1 bj2 . . . bjn .
j1 =0 j2 =0 jn =0

We can use the commutation relation for Bose operators repeatedly to

rearrange the operators b, b† so that
∞ X
X ∞
f (b, b† ) = (n) † m n
fmn (b ) b .
m=0 n=0

We say that the function f (b, b† ) is in normal order form.

(i) Consider the functions

f (b, b† ) = b† bb† b, g(b, b† ) = b† bb† bb† b.

Find the normal order form for these functions.

(ii) Consider the operator exp(−b† b), where  is a real positive parameter.
Find the normal order form.

Solution 21. (i) From the commutation relations for Bose operators we
find bb† = I + b† b. Thus

f (b, b† ) = b† b + b† b† bb, g(b, b† ) = b† b† b† bbb + 3b† b† bb + b† b.

(ii) Using the results from (i) and b† b(b† b)j ≡ b† (I + b† b)j b we find
∞
† X 1 −
e−b b
= (e − 1)j (b† )j bj .
j=0
j!

Problem 22. Let { |ni : n = 0, 1, 2, . . . } be the number states. We

define the linear operators
∞
X ∞
X
Ê := |nihn + 1|, Ê † := |n + 1ihn|.
n=0 n=0

Obviously, Ê † follows from Ê.

(i) Find Ê Ê † and Ê † Ê.
(ii) Let f be an analytic function. Calculate Êf (n̂)Ê † and Ê † f (n̂ + I)Ê,
where n̂ is the photon number operator and I is the identity operator.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 399

Bose Operators and Number States 399

Solution 22. (i) Using hm|ni = δmn and the completeness relation we
find
∞
X ∞
X ∞ X
X ∞
Ê Ê † = |mihm + 1| |n + 1ihn| = |miδm+1,n+1 hn|
m=0 n=0 m=0 n=0
X∞
= |mihm| = I.
m=0

Analogously, we find Ê † Ê = I − |0ih0|.

(ii) Using the Taylor expansion around 0 of an analytic function we have
∞ ∞
X X f (j) (0)
Êf (n̂) = |mihm + 1| n̂j .
m=0 j=0
j!

Applying n̂|mi = m|mi, n̂j |mi = mj |mi and

(n̂ + I)|mi = (m + 1)|mi, (n̂ + I)j |mi = (m + 1)j |mi

we obtain Êf (n̂) = f (n̂ + I)Ê. Thus Êf (n̂)Ê † = f (n̂ + I). Analogously

Ê † f (n̂ + I)Ê = f (n̂).

Problem 23. Consider the Susskind-Glogower canonical phase states

∞
X
|φi := eiφn |ni
n=0

where |ni are the number states. Let

∞
X
L̂ := |n − 1ihn|
n=1

be the non unitary number-lowering operator. Find L̂|φi.

Solution 23. Since hm|ni = δmn we have L̂|φi = eiφ |φi. This means
that |φi is an eigenstate of the operator L̂.

Problem 24. Let b† and b be Bose creation and annihilation operators,

respectively. Consider the operator
2
+α2 (b† )2 +α3 (bb† +b† b)
eα1 b (1)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 400

400 Problems and Solutions

where α1 , α2 , α3 ∈ R. Let  ∈ R be an arbitrary real parameter. Find the

smooth functions f0 , f1 , f2 and f3 , depending on , such that
2
+α2 (b† )2 +α3 (bb† +b† b)) † 2 † 2
e(α1 b = ef0 ()I+f1 ()(b )
ef2 ()b b ef3 ()b (2)
where I denotes the identity operator. Then set  = 1. Solve the problem
using parameter differentiation with respect to . We find a system of
ordinary differential equations for the functions f0 , f1 , f2 and f3 .

Solution 24. Differentiating the left-hand side of (2) with respect to 

yields
† 2 †
2
+b† b))
(α1 b2 + α2 (b† )2 + α3 (bb† + b† b))e(α1 b +α2 (b ) +α3 (bb =
 
f0 ()I+f1 ()(b† )2 df0 df1 † 2 f2 ()b† b f3 ()b2
e I+ (b ) e e
d d
 
† 2 † df2 † 2
+ef0 ()I+f1 ()(b ) ef2 ()b b b b ef3 ()b
d
 
f0 ()I+f1 ()(b† )2 f2 ()b† b f3 ()b2 df3 2
+e e e b .
d

Owing to the identity eX̂ e−X̂ = I, we have

2
+α2 (b† )2 +α3 (bb† +b† b)) 2 † † 2
e−(α1 b = e−f3 ()b e−f2 ()b b e−f0 ()I−f1 ()(b )
.
From the last two equation we obtain
df0 df1 −f3 b2 −f2 b† b † 2 f2 b† b f3 b2
α1 b2 + α2 (b† )2 + α3 (bb† + b† b) = I+ e e (b ) e e
d d
df2 −f3 b2 † f3 b2 df3 2
+ e b be + b .
d d
Since
† †
e−f2 ()b b (b† )2 ef2 ()b b = (b† )2 e−2f2 ()
2 2
e−f3 ()b (b† )2 ef3 ()b = (b† )2 + 4f32 ()b2 − 2f3 ()(I + 2b† b)
2 2
e−f3 ()b b† bef3 ()b = b† b − 2f3 ()b2
we find
α1 b2 + α2 (b† )2 + 2α3 b† b + α3 I
df0 df1 −2f2 † 2 df2 † df3 2
= I+ e ((b ) + 4f32 b2 − f3 (2I + 4b† b)) + (b b − 2f3 b2 ) + b
d d d d
where we used
1 2
e−µÂ B̂eµÂ = B̂ − µ[Â, B̂] + µ [Â, [Â, B̂]] + · · ·
2!
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 401

Bose Operators and Number States 401

and bb† = I + b† b. Separating out terms with I, b2 , (b† )2 , and b† b we find

df0 df1 −2f2
− α3 − 2f3 e =0
d d
df1 −2f2 df3 df2
4f32 e + −2 f3 − α1 = 0
d d d
df1
− α2 e2f2 = 0
d
df1 df2
4 f3 e−2f2 − + 2α3 = 0.
d d
Using this system of equations we can cast the system of nonlinear differ-
ential equations in the form
df0 df1
= α3 + 2α2 f3 , = α2 e2f2
d d
df2 df3
= 2α3 + 4α2 f3 , = α1 + 4α3 f3 + 4α2 f32
d d
with the initial conditions fj (0) = 0 for j = 0, 1, 2, 3. We first solve the
fourth equation which is a Riccati equation and then insert it into the
third and second equation to find f2 and f0 . Finally we solve for f1 . The
integration yields
1
f0 () = ln(cosh(2λ) − (α3 /λ) sinh(2λ))
2
(α2 /2λ) sinh(2λ)
f1 () =
cosh(2λ) − (α3 /λ) sinh(2λ)
f2 () = − ln(cosh(2λ) − (α3 /λ) sinh(2λ))
(α1 /2λ) sinh(2λ)
f3 () =
cosh(2λ) − (α3 /λ) sinh(2λ)
√
where λ := α3 − α1 α2 . Setting  = 1 we have
2
+α2 (b† )2 +α3 (bb† +b† b) 1
eα1 b =p
cosh(2λ) − (α3 /λ) sinh(2λ)
 
(α2 /λ) sinh(2λ) † 2
× exp (b )
cosh(2λ) − (α3 /λ) sinh(2λ)
× exp(ln(cosh(2λ) − (α3 /λ) sinh(2λ))−1 b† b)
 
(α1 /2λ) sinh(2λ) 2
× exp b .
cosh(2λ) − (α3 /λ) sinh(2λ)

Problem 25. The homogeneous Bogolubov transform of the Bose creation

operator b† and Bose annihilation operator b
eb = µb + νb† , µ, ν ∈ C
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 402

402 Problems and Solutions

for a pair of complex parameters

µ = |µ| exp(iφ), ν = |ν| exp(iθ)
obeying additionally |µ|2 − |ν|2 = 1 is canonical since it leaves the commu-
tator invariant, i.e.
[b, b† ] = [eb, eb† ] = I.
Every canonical transform can be represented as a unitary transformation
eb = B(µ, ν)bB † (µ, ν).

The Bogolubov unitary operator B(µ, ν) is defined by this relation up to an

arbitrary phase factor. One choice is the normal form
   ∗ 
−1/2 ν †2 †

ν 2
B(µ, ν) = µ exp − b exp − ln(µ)b b exp b .
2µ 2µ
Show that the Bogolubov transform forms a continuous non-commutative
group.

Solution 25. Let |µ0 |2 − |ν 0 |2 = 1, |µ00 |2 − |ν 00 |2 = 1. Then we have

B(µ0 , ν 0 )B(µ00 , ν 00 ) = B(µ, ν)
and µ = µ0 µ00 + ν 0∗ ν 00 , ν = µ0∗ ν 00 + ν 0 µ00 with |µ|2 − |ν|2 = 1. The identity
element of the group is given by B(1, 0), where we used that ln(1) = 0. The
inverse element of B(µ, ν) is given by
B −1 (µ, ν) = B † (µ, ν) = B(µ∗ , −ν).
Obviously, the associative law also holds.

Problem 26. Consider the operators

1 †
K+ := b†1 b†2 , K− := b2 b1 , K3 := (b b1 + b†2 b2 + I)
2 1
where I is the identity operator. Find the commutators
[K+ , K− ], [K3 , K+ ], [K3 , K− ].
The operators K+ , K− , K3 form a representation of the Lie algebra su(1, 1).

Solution 26. Using the commutation relations given above we find

[K+ , K− ] = b†1 b†2 b2 b1 − b2 b1 b†1 b†2 = b†1 b†2 b2 b1 − b2 b†2 − b†1 b1 b2 b†2
= −b2 b†2 − b†1 b1 = −I − b†2 b2 − b†1 b1
= −2K3 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 403

Bose Operators and Number States 403

Analogously [K3 , K− ] = −K− , [K3 , K+ ] = K+ .

Problem 27. Consider the linear operators

1 †
J+ := b†1 b2 , J− := b†2 b1 , J3 := (b b1 − b†2 b2 )
2 1

where b†1 , b†2 are Bose creation operators and b1 , b2 are Bose annihilation
operators and I is the identity operator. Find the commutators

[J+ , J− ], [J3 , J+ ], [J3 , J− ].

The operators J+ , J− , J3 form a representation of the semisimple Lie alge-

bra su(2).

Solution 27. Using the commutation relation given above we find

[J+ , J− ] = b†1 b2 b†2 b1 − b†2 b1 b†1 b2 = b†1 b2 b†2 b1 − b†2 b2 − b†2 b†1 b1 b2
= −b†2 b2 + b†1 b1 = 2J3 .

Analogously [J3 , J− ] = −J− , [J3 , J+ ] = J+ .

Problem 28. Consider the Bose creation operators b†1 , b†2 and Bose an-
nihilation operators b1 , b2 with b1 = b ⊗ I, b2 = I ⊗ b and b1 |00i = 0|00i,
b2 |00i = 0|00i, where |00i ≡ |0i ⊗ |0i is the vacuum state. Consider the
linear transformation
eb1 = u11 b1 + u12 b2 + v11 b† + v12 b†
1 2
eb2 = u21 b1 + u22 b2 + v21 b† + v22 b†
1 2
eb† = v ∗ b1 + v ∗ b2 + u∗ b† + u∗ b†
1 11 12 11 1 12 2
eb† = v ∗ b1 + v ∗ b2 + u∗ b† + u∗ b†
2 21 22 21 1 22 2

where ujk , vjk ∈ C.

(i) Find the condition that the operators eb1 , eb2 , eb†1 , eb†2 also satisfy the
commutation relation for Bose operators.
(ii) For the vacuum state of the Bose fields eb1 , eb2 we can write
∞ X
X ∞
|0i
e ≡ |e0i ≡ |e
0e 0i ⊗ |e
0i = cmn |mi ⊗ |ni.
m=0 n=0

Find the recurrence relation for cmn from the condition

eb1 |e
0i ⊗ |e
0i = 0, eb2 |e
0i ⊗ |e
0i = 0. (1)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 404

404 Problems and Solutions

Solution 28. (i) From the conditions

[eb1 , eb2 ] = 0, [eb1 , eb†2 ] = 0, [eb1 , eb†1 ] = I, [eb2 , eb†2 ] = I

we find
u11 v21 + u12 v22 − u21 v11 − u22 v12 = 0 (2a)
u11 u∗21 + u12 u∗22 − v11 v21
∗ ∗
− v12 v22 =0 (2b)
u11 u∗11 + u12 u∗12 − ∗
v11 v11 − ∗
v12 v12 =1 (2c)
u21 u∗21 + u22 u∗22 − v21 v21
∗ ∗
− v22 v22 = 1. (2d)
(ii) From the conditions (1) we find
√ √
c(m+1)n u11 m + 1 + cm(n+1) u12 n + 1
√ √
+c(m−1)n v11 m + cm(n−1) v12 n = 0 (3a)
and √ √
c(m+1)n u21 m + 1 + cm(n+1) u22 n + 1
√ √
+c(m−1)n v21 m + cm(n−1) v22 n = 0. (3b)
Let

∆1 := u11 u22 − u12 u21 , ∆2 := u11 v21 − u21 v11 , ∆3 := u11 v22 − u21 v12

∆4 := u22 v11 − u12 v21 , ∆5 := u22 v12 − u12 v22 .

Multiplication of (3a) with u21 and (3b) with u11 and subtracting yields
√ √ √
cm(n+1) ∆1 n + 1 = −c(m−1)n ∆2 m − cm(n−1) ∆3 n.

Multiplication of (3a) with u22 and (3b) with u12 and subtracting yields
√ √ √
c(m+1)n ∆1 m + 1 = −c(m−1)n ∆4 m − cm(n−1) ∆5 n.

We assumed that ∆1 6= 0. From (2a) we see that ∆2 = ∆5 . Thus we have

c(2k)(2n+1) = c(2k+1)(2n) = 0

p p
c(2k)(2n) = (−1)n+k (2n)! (2k)!
X  ∆2 2s  ∆3 n−s  ∆4 k−s 1
× c0
∆1 2∆1 2∆1 (n − s)!(k − s)!(2s)!
0≤s≤n
s≤k

and p p
c(2k+1)(2n+1) = (−1)n+k+1 (2n + 1)! (2k + 1)!
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 405

Bose Operators and Number States 405

X  ∆2 2s+1  ∆3 n−s  ∆4 k−s 1

× c0 .
∆1 2∆1 2∆1 (n − s)!(k − s)!(2s + 1)!
0≤s≤n
s≤k

Consequently, for the vacuum state of Bose operators eb1 and eb2 we find
∞
X

|e
0e
0i = c(2k)(2n) |2ki ⊗ |2ni + c(2k+1)(2n+1) |2k + 1i ⊗ |2n + 1i .
k=0,n=0

In operator form this can be written as

 
∆4 † 2 ∆2 † † ∆3 † 2
|e0i = c0 exp −
0e (b1 ) − b1 b2 − (b2 ) |0i ⊗ |0i.
2∆1 ∆1 2∆1
Thus the unitary operator
 
∆4 † 2 ∆2 † † ∆3 † 2
U = exp − (b1 ) − b1 b2 − (b2 )
2∆1 ∆1 2∆1
is the operator of transformation of the vacuum states for the most general
two-dimensional Bogolubov transformation. Thus we also have

|mi
e ⊗ |e
ni = U (|mi ⊗ |ni).

Problem 29. Quantum mechanically, a phase shift δ induced by a linear

optical element on a single-mode optical field is described by the unitary
operator
U := exp(iδn̂)
where n̂ := b† b is the number operator and b the annihilation operator for
the optical mode. Assume the optical field is in the state |ψi.
(i) Express |ψi in the basis of photon number state representation.
(ii) Find the state |ψ 0 i := U |ψi.
(iii) Find |∆ψi := |ψ 0 i − |ψi and the norm k |∆ψik.

Solution 29. (i) We can write

∞
X
|ψi = cm |mi
m=0

in the basis of photon number state representation, where cm are the ex-
pansion coefficients.
(ii) The phase-shifted state |ψ 0 i can be written as
∞
X ∞
X
|ψ 0 i = exp(iδn̂) cm |mi = cm eiδm |mi.
m=0 m=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 406

406 Problems and Solutions

(iii) Thus for the difference we find

∞
X
|∆ψi = |ψ 0 i − |ψi = cm (eiδm − 1)|mi
m=0

and therefore
∞
X ∞
X
k |∆ψik2 = h∆ψ|∆ψi = 4 |cm |2 sin2 (δm/2) = 4 Pm sin2 (δm/2)
m=0 m=0

where Pm = |cm |2 is the photon number distribution for the input field.

Problem 30. The generator of displacements for numbers is formally

defined by Z π
D(k) := dφ eikφ |φihφ|
−π
where
∞
X
|φi = einφ |ni, φ ∈ R.
n=0

Show that these basis states are not normalized.

Solution 30. Since

∞
X
hφ| = hm|e−imφ
m=0

and hm|ni = δmn we find

∞
X
hφ|φi = 1.
m=0

Problem 31. Let b1 , b2 be Bose annihilation operators. Show that

† † 1 † †
eµb1 b2 eνb1 b2 |00i = eνb1 b2 /(1−µν) |00i, µ, ν ∈ R (1)
1 − µν
where |00i ≡ |0i ⊗ |0i.

Solution 31. We solve the problem by considering the expression

† † † †
eµb1 b2 eνb1 b2 |00i = ef (µ,b1 ,b2 ) |00i

where f is an analytic function. Differentiating both sides with respect to

µ yields
† † ∂f
b1 b2 eµb1 b2 eνb1 b2 |00i = ef |00i.
∂µ
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 407

Bose Operators and Number States 407

Thus
∂f
b1 b2 ef |00i = ef |00i.
∂µ
Note that ∂f /∂µ commutes with exp(f ) since f is a function of b†1 and b†2
only. If we multiply from the left by exp(−f ) we find

∂f
e−f b1 b2 ef |00i = |00i.
∂µ
If follows that
∂f
e−f b1 ef e−f b2 ef |00i = |00i.
∂µ
Using
∂g
[b, g(b, b† )] = (2)
∂b†
with g = ef , we obtain
!
−f f −f f ∂ef ∂f
e b1 e = e e b1 + = b1 +
∂b†1 ∂b†1

since ef commutes with ∂f /∂b†1 . Similarly

∂f
e−f b2 ef = b2 + .
∂b†2

Thus we have
! !
∂f ∂f ∂f
b1 + b2 + |00i = |00i.
∂b†1 ∂b†2 ∂µ

Since b2 |00i = 0|00i we arrive at

!
∂f ∂f ∂f ∂f
b1 + |00i = |00i.
∂b†2 ∂b†1 ∂b†2 ∂µ

Using (2) again with g = ∂f /∂b†2 we obtain

∂f ∂f ∂2f
b1 = b1 + .
∂b†2 ∂b†2 ∂b†1 ∂b†2

Since b1 |00i = 0|00i we obtain

!
∂2f ∂f ∂f ∂f
+ |00i = |00i.
∂b†1 ∂b†2 ∂b†1 ∂b†2 ∂µ
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 408

408 Problems and Solutions

Since f contains only b†1 and b†2 which commute, the solution of this partial
differential equation must be of the form

f (µ, b†1 , b†2 ) = h1 (µ)I + h2 (µ)b†1 b†2 .

Thus f (0, b†1 , b†2 ) = νb†1 b†2 or h1 (0) = 0, h2 (0) = ν owing to (1). Inserting
this ansatz into the partial differential equation and equating equal powers
of b†1 b†2 , we find that h1 and h2 satisfy the nonlinear system of ordinary
differential equations

dh2 dh1
= h22 , = h2
dµ dµ

with the solution of the initial value problem

ν
h2 (µ) = , h1 (µ) = − ln(1 − µν)
1 − µν

and thus we find (1).

Problem 32. The standard Pauli group for continuous variable quantum
computing of n coupled oscillator systems is the Heisenberg-Weyl group
which consists of phase-space displacement operators for n harmonic os-
cillators. This group is a continuous Lie group and can therefore only be
generated by a set of continuously parameterized operators. The Lie alge-
bra that generates this group is spanned by the 2n canonical operators p̂j ,
q̂j , j = 1, 2, . . . , n along with the commutation relation

[q̂j , p̂k ] = i~δjk I.

For a single oscillator (n = 1) the algebra is spanned by the canonical

operators {q̂, p̂, I}. We define the operators

X(q) := e−(i/~)qp̂ , Z(p) := e(i/~)pq̂

where q, p ∈ R. Let { |si : s ∈ R} be position eigenstates (in the sense of

generalized functions).
(i) Calculate the states X(q)|si, Z(p)|si.
(ii) Find the commutator [X(q), Z(p)].

Solution 32. (i) We find in the sense of generalized functions that

X(q)|si = |s + qi, Z(p)|si = exp((i/~)ps)|si.

Thus the operator X(q) is a position translation operator. The operator

Z(p) is a momentum boost operator.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 409

Bose Operators and Number States 409

(ii) We obtain X(q)Z(p) = e−(i/~)qp Z(p)X(q) and

[X(q), Z(p)] = (I − e(i/~)qp )X(q)Z(p).

Problem 33. Let r ∈ R. Find 1 , 2 and 3 such that

† † † † † †
er(b1 b2 −b1 b2 ) ≡ e1 b1 b2 e2 (b1 b1 +b2 b2 +I) e3 b1 b2 .

Solution 33. Using the fact that the operators

1
K+ := −b1 b2 , K− := b†1 b†2 , K3 := − (b†1 b1 + b†2 b2 + I)
2
form the semi-simple Lie algebra

[K3 , K+ ] = K+ , [K3 , K− ] = −K− , [K+ , K− ] = 2K3

and
er(K+ +K− ) ≡ eK− tanh(r) e2 ln(cosh(r))K3 eK+ tanh(r)
we find 1 = tanh(r), 2 = − ln(cosh(r)), 3 = − tanh(r).

Problem 34. To build a simple quantum computer one could use the
following optical gates

US := exp(iπb† b) phase modulator

π 
UB := exp (b†1 b2 − b1 b†2 ) quantum beam splitter
4
χ 
UF := exp b†3 b3 (b†1 b2 − b1 b†2 ) Fredkin gate
2
(i) Calculate the state US |ni.
(ii) Calculate the state UB |01i.
(iii) Calculate the states UF |011i, UF |101i, UF |xy0i with χ = π and x, y ∈
{ 0, 1 }.

Solution 34. (i) Since b† b|ni = n|ni we obtain

†
eiπb b |ni = eiπn |ni = (−1)n |ni.

(ii) Since

(b†1 b2 − b1 b†2 )|01i = |10i, (b†1 b2 − b1 b†2 )|10i = −|01i

we find
1
UB |01i = √ (|01i + |10i)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 410

410 Problems and Solutions

√ √
where we used sin(π/4) = 1/ 2 and cos(π/4) = 1/ 2.
(iii) Since

b†3 b3 (b†1 b2 − b1 b†2 )|011i = |101i, b†3 b3 (b†1 b2 − b1 b†2 )|101i = −|011i

and b†3 b3 (b†1 b2 − b1 b†2 )|xy0i = 0|xy0i we find the states

UF |101i = −|011i, UF |011i = |101i, UF |xy0i = |xy0i

where we used that b|0i = 0|0i and b|1i = |0i. Thus b†3 b3 plays the role of
a control operator.

Problem 35. Let  be a real parameter, σ3 the Pauli spin matrix and I2
the 2 × 2 identity matrix. Calculate
† †
f () = eσ3 ⊗(b−b ) (I2 ⊗ b† b)e−σ3 ⊗(b−b )
(1)

using parameter differentiation and then solving the differential equation

with the corresponding initial values (operators), i.e. f (0) = I2 ⊗ b† b.

Solution 35. From (1) we obtain f (0) = I2 ⊗ b† b. Now

df † †
= eσ3 ⊗(b−b ) (σ3 ⊗ ((b − b† )b† b − b† b(b − b† )))e−σ3 ⊗(b−b )
d
† †
= eσ3 ⊗(b−b ) (σ3 ⊗ (b + b† ))e−σ3 ⊗(b−b ) .

Thus
df (0)
= σ3 ⊗ (b + b† ).
d
Since (b − b† )(b + b† ) − (b + b† )(b − b† ) = 2I and σ32 = I2 we obtain for the
second order derivative
d2 f
= 2I2 ⊗ I.
d2
Thus the solution of this second order linear differential equation is

f () = 2 I2 ⊗ I + C1  + C2 .

Inserting the two initial values (operators) yields

C1 = σ3 ⊗ (b + b† ), C2 = I2 ⊗ b† b.

It follows that

f () = 2 I2 ⊗ I + σ3 ⊗ (b + b† ) + I2 ⊗ b† b.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 411

Bose Operators and Number States 411

Problem 36. Let b, b† be Bose operators. Find the eigenvalues of the

† †
operator (−1)b b . This unitary operator is defined by eiπb b . Why is this
operator called the parity operator?

Solution 36. Let |ni be a number state. Then b† b|ni = n|ni, b† bb† b|ni =
n2 |ni etc. Thus
†
eiπb b |ni = eiπn |ni
where n = 0, 1, 2, . . . , ∞. It follows that

hm|eiπn |ni = eiπn hm|ni = eiπn δmn .

Since eiπn = +1 if n is even and eiπn = −1 if n is odd we find that the

†
eigenvalues of (−1)b b are +1 and −1 both infinitely degenerate.

Problem 37. The hermitian cosine ĈSG and sine ŜSG operators intro-
duced by Susskind and Glogower are given in the number state basis by
∞
1X
ĈSG := (|nihn + 1| + |n + 1ihn|)
2 n=0
∞
1 X
ŜSG := (|nihn + 1| − |n + 1ihn|).
2i n=0

Solve the eigenvalue equations ĈSG |ciSG = c|ciSG , ŜSG |siSG = s|siSG .

Solution 37. We find

r ∞ r ∞
2p X 2p X
in Un (s)|ni
4 4
|ciSG = 1 − c2 Un (c)|ni, |siSG = 1 − s2
π n=0
π n=0

where Un (x) for x = c or x = s are the Chebyshev polynomials of second

kind and x ∈ [−1, 1]. They obey the recursion formula

Un+1 (x) − 2xUn (x) + Un−1 (x) = 0

where U0 (x) = 1 and U1 (x) = 2x.

Problem 38. Let |0i be the vacuum state, i.e. b|0i = 0|0i, and  ∈ R.
Calculate the state
exp(b† ⊗ b† )(|0i ⊗ |0i).

Sometimes one also writes exp(b†1 b†2 )|0i|0i.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 412

412 Problems and Solutions

√
Solution 38. Since b† |ni = n + 1|n + 1i for n = 0, 1, 2, . . . we obtain
(b† ⊗ b† )(|ni ⊗ |ni = (b† |ni) ⊗ (b† |ni) = (n + 1)|n + 1i ⊗ |n + 1i.
Now
 † 2
exp(b† ⊗ b† )|0i ⊗ |0i = (I ⊗ I + b ⊗ b† + (b† )2 ⊗ (b† )2 + · · ·)|0i ⊗ |0i
1! 2!
= |0i ⊗ |0i + |1i ⊗ |1i + 2 |2i ⊗ |2i + · · ·
X∞
= j |ji ⊗ |ji.
j=0

What is the condition on  so that the series converges? This means: what
is the condition that the state can be normalized?

Problem 39. Let b, b† be Bose annihilation and creation operators, re-

spectively. Let  ∈ R and |0i be the vacuum state, i.e. b|0i = 0|0i.
(i) Calculate the state b exp(b† )|0i.
(ii) Calculate the state bn exp(b† )|0i, (n = 2, 3, . . .).

Solution 39. (i) Since

†
bb |0i = |0i, b(b† )2 |0i = 2b† |0i, b(b† )3 |0i = 3(b† )2 |0i
and in general b(b† )n |0i = n(b† )n−1 |0i we obtain
2 † 3
b exp(b† )|0i = |0i + 2b |0i + 3(b† )2 |0i + · · ·
2! 3!
2

= (I + b† + (b† )2 + · · ·)|0i
2!
=  exp(b† )|0i.
(ii) Using the result from (i) we have
bn exp(b† )|0i = n exp(b† )|0i.

Problem 40. Let |0i be the vacuum state. Calculate the state
1 †
(b ⊗ I − I ⊗ b† )(b† ⊗ I + I ⊗ b† )(|0i ⊗ |0i)
2
where I is the identity operator. Discuss.

Solution 40. Since (b† ⊗ I)(I ⊗ b† ) = (I ⊗ b† )(b† ⊗ I) we obtain the state

1
√ (|2i ⊗ |0i − |0i ⊗ |2i) .
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 413

Bose Operators and Number States 413

The photons interfere constructively or destructively. Complete destructive

interference implies that the affected outgoing mode is in the vacuum state.

Problem 41. Consider the operators

1 † 1 † 1 †
Jx = (b b2 + b†2 b1 ), Jy = i(b b2 − b†2 b1 ), Jz = (b b1 − b†2 b2 )
2 1 2 1 2 1

and J+ := Jx + iJy = b†1 b2 , J− := Jx − iJy = b†2 b1 . Let φ ∈ R. Find

e−iJz φ b†1 eiJz φ , e−iJz φ b†2 eiJz φ , e−iJy φ b†1 eiJy φ , e−iJy φ b†2 eiJy φ .

Solution 41. We obtain

exp(−iJz φ)b†1 exp(iJz φ) = b†1 e−iφ/2

exp(−iJz φ)b†2 exp(iJz φ) = b†2 e−iφ/2
exp(−iJy φ)b†1 exp(iJy φ) = b†1 cos(φ/2) + b†2 sin(φ/2)
exp(−iJy φ)b†2 exp(iJy φ) = b†2 cos(φ/2) − b†1 sin(φ/2).

Problem 42. Let { |ni : n = 0, 1, 2, . . . } be number states. Consider the

linear operator
∞
X
T13 := (|ni ⊗ I ⊗ I)(I ⊗ I ⊗ hn|)
n=0

in the product (infinite-dimensional) Hilbert space H = H1 ⊗ H2 ⊗ H3 with

H1 = H2 = H3 . Here I denotes the identity operator. Apply the operator
T13 to the state I ⊗ I ⊗ |ψi.

Solution 42. We find

∞
X
T13 (I ⊗ I ⊗ |ψi) = (|ni ⊗ I ⊗ I)(I ⊗ I ⊗ hn|ψiI)
n=0
X∞
= (hn|ψi|ni) ⊗ I ⊗ I
n=0
= |ψi ⊗ I ⊗ I

where we used
∞
X
|ψi = hn|ψi|ni.
n=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 414

414 Problems and Solutions

The operator T13 can be considered as a transfer operator.

Problem 43. A beam splitter can be realized by means of a linear medium

where the polarization vector is proportional to the incoming electric field

P̂ = χÊ

with χ ≡ χ(1) denoting the first order (linear) susceptibility. We consider

the incoming field excited only in the relevant spatial modes b1 and b2 (at
the same frequency ω)
r
~ω  
Ê(r, t) = i (b1 + b2 )ei(k·r−ωt) + h.c.
20 V
where h.c. denotes the hermitian conjugate. The interaction Hamilton op-
erator contains only the resonant terms
χ~ω †
ĤI = −P̂ · Ê = −χÊ2 = (b b2 + b1 b†2 )
20 V 1
where · denotes the scalar product. The evolution operator (in the interac-
tion picture) of the whole device is expressed as
r ! !
1−τ  † †

U := exp i arctan b1 b2 + b1 b2
τ

where τ , given by
  −1
χ~ω
τ= 1 + tan2
20 V
represents the transmissivity of the beam splitter.
(i) Calculate eb1 = U † b1 U , eb2 = U † b2 U .
(ii) Find a rotation of the phase frame by 3π/2.

Solution 43. (i) Straightforward calculation yields

eb1 = U † b1 U = −iτ 1/2 b1 + (1 − τ )1/2 b2

eb2 = U † b2 U = i(1 − τ )1/2 b1 + τ 1/2 b2 .

(ii) A rotation of the phase frame can be obtained by the substitution
b1 → −ib1 . Then we obtain
eb1 = τ 1/2 b1 + (1 − τ )1/2 b2 , eb2 = τ 1/2 b2 − (1 − τ )1/2 b1 .

Problem 44. Owing to their helical wave fronts the electromagnetic field
of photons having an orbital angular momentum has a phase singularity.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 415

Bose Operators and Number States 415

There the intensity has to vanish resulting in a doughnut-like intensity

distribution. These light fields can be described using Laguerre Gaussian
(LGpl ) modes with two indices p and l. The p-index (p = 0, 1, 2, . . .) iden-
tifies the non-axial radial nodes observed in the transversal plane and the
l-index (l = 0, ±1, ±2, . . .) the number of the 2π-phase shifts along a closed
path around the beam center. The index l is also called the topological
winding number. It describes the helical structure of the wave front around
a wave front singularity or dislocation. The index l also determines the
amount of orbital angular momentum in units of ~ carried by one pho-
ton. When the pump beam is a LGl0 p0 mode, under conditions of collinear
phase-matching, the two-photon state at the output of the nonlinear crys-
tal can be written as a coherent superposition of eigenstates of the orbital
angular-momentum operator that are correlated in orbital angular momen-
tum, i.e., l1 +l2 = l0 , where l1 and l2 refer to the orbital angular momentum
eigenvalues for the signal and idler photons. A photon state described by
a LG mode can be written as
Z
|lpi := dqLGlp (q)b† (q)|0i

where the mode function in the spatial frequency domain is given by

 1/2  |l|  2 2  2 2
ω0 p! ω0 ρk ρk ω0 ρ ω
LGlp (ρ, φ) := √ L|l|
p exp − k 0
2π(|l| + p)! 2 2 4
   
|l|
× exp ilφk + i p − π
2

with ρk and φk being the modulus and phase, respectively, of the transverse
|l|
coordinate q. The functions Lp are the associated Laguerre polynomials
and ω0 is the beam width. Find the state |lpi for l = p = 0.

Solution 44. Since the associated Laguerre polynomial L00 is given by

L00 (x) = 1 we obtain
 ω 1/2  2 2
0 ρ ω
LG00 = exp − k 0 .
2π 4

Thus for LG00 we find a Gaussian.

Programming Problem

Problem 1. Find the normal ordering of bb† bb† and apply it then to the
vacuum state. Find the normal ordering of bb† b† and apply it then to the
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 416

416 Problems and Solutions

vacuum state. Find (b + b† )4 . Find the operator b + bb† + bb† b† + bb† b† b†

and normal ordering. Apply computer algebra.

Solution 1. In the SymbolicC++ program the operator b is denoted by b,

the operator b† by bd and the vacuum state by vs. The rules b*bd==bd*b+1
and b*vs=0 are implemented.

// bose2.cpp
#include <iostream>
#include "symbolicc++.h"
using namespace std;

int main(void)
{
Symbolic b("b"), bd("bd"), vs("vs");
b = ~b; bd = ~bd; vs = ~vs; // noncommutative
Equations rules = (b*bd==bd*b+1,b*vs==0);
// example 1
Symbolic res1 = b*bd*b*bd;
cout << "res1 = " << res1.subst_all(rules) << endl;
cout << "res1*vs = " << (res1*vs).subst_all(rules) << endl;
// example 2
Symbolic res2 = b*bd*bd;
cout << "res2 = " << res2.subst_all(rules) << endl;
cout << "res2*vs = " << (res2*vs).subst_all(rules) << endl;
// example 3
Symbolic res3 = (b+bd)^4;
cout << "res3 = " << res3.subst_all(rules) << endl;
cout << "res3*vs = " << (res3*vs).subst_all(rules) << endl;
// example 4
Symbolic res4 = b + b*bd + b*bd*bd + b*bd*bd*bd;
cout << "res4 = " << res4.subst_all(rules) << endl;
cout << "res4*vs = " << (res4*vs).subst_all(rules) << endl;
return 0;
}

The output is

res1 = bd^(2)*b^(2)+3*bd*b+1
res1*vs = vs
res2 = bd^(2)*b+2*bd
res2*vs = 2*bd*vs
res3 = b^(4)+4*bd*b^(3)+6*b^(2)+6*bd^(2)*b^(2)
+12*bd*b+4*bd^(3)*b+6*bd^(2)+bd^(4)+3
res3*vs = 3*vs+6*bd^(2)*vs+bd^(4)*vs
res4 = b+bd*b+bd^(2)*b+2*bd+bd^(3)*b+3*bd^(2)+1
res4*vs = vs+2*bd*vs+3*bd^(2)*vs
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 417

Bose Operators and Number States 417

18.3 Supplementary Problems

Problem 1. (i) Let b, b† be Bose creation and annihilation operators and

1 †
Ĥ = (b b + bb† ).
2

Show that [b, Ĥ] = b.

(ii) Let n be a positive integer. Show that

[b† b, (b† )n ] = n(b† )n , [b† b, bn ] = −nbn .

(iii) Let b† , b be Bose creation and annihilation operators and z ∈ C. Show

that
† † † †
[b, e−zb b ] = (ez − 1)e−zb b b, [b† , e−zb b ] = (ez − 1)e−zb b b† .

Problem 2. Let f : C → C be an analytic function.

(i) Show that ezb f (b† )|0i = f (b† + zI)|0i.
†
(ii) Show that e−zb b f (b† )|0i = f (b† e−z )|0i.

Problem 3. Let f, g : C → C be analytic functions. Then

∞
X 1 dj dj
[f (b), f (b† )] = f (b).
j=1
j! d(b† )j dbj

Let τ ∈ R. Use this relation to calculate the commutators

[b, exp(τ (b† )2 /2)], [b2 , exp(τ (b† )2 /2], [b3 , exp(τ (b† )2 /2].

Then show that b exp(τ (b† )2 /2)|0i = τ b† exp(τ (b† )2 /2)|0i.

Problem 4. Let b† , b be Bose creation and annihilation operators. Show

that
∞
X (eθ − 1)j
exp(θb† b) = (b† )j bj .
j=0
j!

Problem 5. Let b†1 , b†2 , b1 , b2 be Bose creation and annihilation operators

and let N ≥ 1 be a positive integer. Consider the operators
√ √
2 † † 2i
T̂1 = 3/2 (b1 b1 b2 + b2 b1 b1 ), T̂2 = 3/2 (b†1 b†1 b2 − b†2 b1 b1 ),
†
N N
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 418

418 Problems and Solutions

2
(2b†2 b2 − b†1 b1 ).
T̂3 =
N
Show that the commutators are given by
i 4i
(I − T̂3 )(I + 3T̂3 ) + 2 I,
[T̂1 , T̂2 ] =
N N
4i 4i
[T̂3 , T̂1 ] = T̂2 , [T̂3 , T̂2 ] = − T̂1
N N
where I denotes the identity operator.

Problem 6. Let b† , b be Bose creation and annihilation operators. Show

that †
ρ(b, b† ) = (1 − e−λ )e−λb b

is a density operator, where λ ≡ β~ω, β = 1/(kB T ).

Problem 7. Let |ni be a number state and |βi be coherent state. Show
that
(|β|2 )n
|hn|βi|2 = exp(−|β|2 )
n!
which is a Poisson distribution.

Problem 8. Let b† , b be Bose creation and annihilation operators. Con-

sider the operators
 
1
T̂1 = (b† )2 , T̂2 = −b2 , T̂3 = 4 b† b + I .
2
Show that they satisfy the commutation relations
[T̂1 , T̂2 ] = T3 , [T̂1 , T̂3 ] = −8T1 , [T̂2 , T̂3 ] = 8T2 .
Show that
exp(((b† )2 − b2 )) =
1 1 1
exp(− b2 tanh(2)) exp((b† b + I) ln(cosh(2)) exp( (b† )2 tanh(2)).
2 2 2

Problem 9. The semi-simple Lie algebra su(1, 1) is given by the com-

mutation relations [k1 , k2 ] = −ik3 , [k3 , k1 ] = ik2 , [k2 , k3 ] = −ik1 , where
k1 , k2 and k3 are the basis elements of the Lie algebra. Show that an
infinite-dimensional matrix representation is given by
0 1 0 0 ... 0 i 0 0 ...
   
1 0 2 0 ...  −i 0 2i 0 ...
1 0 2 0 3 ...  , k2 = 1  0 −2i

0 3i . . . 

k1 = 
2 0 0 3 0 ...
  2 0

0 −3i 0 . . . 

.. .. .. .. . . .. .. .. .. . .
. . . . . . . . . .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 419

Bose Operators and Number States 419

1
k3 = diag(1, 3, 5, 7, . . .).
2

Problem 10. Let b† , b be the Bose creation and annihilation operators

and S3 , S+ , S− be the spin- 12 matrices. Consider the operators

N̂ = b† b ⊗ I2 + IB ⊗ S3 , K̂ = (b† + b) ⊗ (S+ + S− ).
(i) Show that [N̂ , b ⊗ S+ + b† ⊗ S− ] = 0B ⊗ 02 .
(ii) Show that [N̂ , K̂] = 2(b† ⊗ S+ − b ⊗ S− ).

Problem 11. Consider the operators

1 † 1
Q+ = (b1 ⊗ c + b2 ⊗ c† ), Q− = (b1 ⊗ c† − b†2 ⊗ c).
2 2
Find Q+ Q− , Q− Q+ and the anticommutator [Q+ , Q− ]+ .

Problem 12. Let

J+ = b†1 b2 ⊗ IF + IB ⊗ c†
J− = b†2 b1 ⊗ IF + IB ⊗ c
J3 = (b†1 b1 − b†2 b2 ) ⊗ IF + IB ⊗ (2c† c − IF ).
Show that [J+ , J− ] = J3 , [J+ , J3 ] = −2J+ , [J− , J3 ] = 2J− .

Problem 13. Find the matrix representation of b† ⊗ c + b ⊗ c† applying

the basis |ni ⊗ |0i, |ni ⊗ c† |0i. Find the matrix representation using the
basis
|βi ⊗ |0i, |βi ⊗ c† |0i
where |βi are coherent states.

Problem 14. Let c† , c be Fermi creation and annihilation operators.

†
(i) Show that eiπc c = IF − 2c† c.
†
(ii) Show that eiπb b = diag(1, −1, 1, −1, . . .).

Problem 15. Find the states

†
b⊗c† c) †
b⊗c† c)
eiπ(b (|ni ⊗ |0i), eiπ(b (|ni ⊗ |0i).

Problem 16. Let b be a Bose operator and I be the identity operator.

Show that we can write
Z Z
1 †
δ(βI − b)δ(βI − b† ) = 2 e−iµ(βI−b) e−iµ(β−b ) d(<(µ))d(=(µ)).
π R R
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 420

420 Problems and Solutions

Problem 17. Let σ1 , σ2 , σ3 be the Pauli spin matrices and b† , b be Bose

creation and annihilation operators. Study the spectrum of the Hamilton
operator

Ĥ = ~ω1 IB ⊗ σ3 + ~ω2 b† b ⊗ I2 + ~ω3 (b† + b) ⊗ σ1 .

Problem 18. Let fˆ(k) be the Fourier transform of f (x), where f ∈

L2 (R) ∩ L1 (R). Show that if

x2 |f (x)|2 dx k |fˆ(k)|2 dk
R R 2
R R
Ex := R and Ek :=
|f (x)|2 dx |fˆ(k)|2 dk
R
R R

exist, then Ex Ek ≥ 1/4.

Problem 19. The Fourier transform is given by

Z
1 ˆ
fˆ(k) = f (x)eikx dx, f (x) = f (k)e−ikx dk.
R 2π

(i) Show that for the Gaussian distribution we have

(x − E)2
   
1 1 2 2
√ exp − ⇔ exp ikE − σ k .
2πσ 2 2σ 2 2

(ii) Show that for Poisson distribution we have

∞
X 1 n
λ exp(−λ)δ(x − n) ⇔ exp(λ(eik − 1))
n=0
n!

(iii) Show that for Lorentzian distribution we have

Γ 1
⇔ exp(imk − |kΛ|).
π (x − m)2 + Γ2

(iv) Show for the product of two Gaussian distributions we have

 
1 2 2 2
exp ik(E1 + E2 ) − (σ1 + σ2 )k
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 421

Chapter 19

Coherent States

19.1 Introduction
Quantum coherent states are the closest quantum-mechanical analogue to
a classical particle oscillating in a harmonic potential. Coherent states are
minimum uncertainty states. Quantum computation circuits with coherent
states as the logical qubits can be constructed using simple linear networks,
conditional measurements and coherent superposition resource states. Co-
herent states are very sensitive to their environment. The output of a single
mode stabilised laser can be described by a coherent state |βi, where β is a
complex number which determines the average field amplitude. Harmonic
oscillator coherent states can be defined in three different equivalent ways.
Firstly, the coherent states are the eigenstates of the Bose annihilation
operator
b|βi = β|βi, β ∈ C.
Thus the spectrum of the operator b fills the entire complex plane. Secondly,
they are displaced vacuum states

|βi = exp(−|β|2 /2) exp(βb† ) exp(−β ∗ b)|0i

where |0i is the vacuum state with h0|0i = 1. Since

b|0i = 0|0i

we have
|βi = exp(−|β|2 /2) exp(βb† )|0i.

421
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 422

422 Problems and Solutions

Thirdly, coherent states are states of minimum uncertainty

~
∆p∆x =
2
and are thus most classical within the quantum framework.

Expressed in number states |ni the coherent states are given by

∞
2 X βn
|βi = e−|β| /2
√ |ni.
n=0 n!

The displacement operator is defined as

D(β) := exp(βb† − β ∗ b).

Then the coherent states can be defined as

|βi = D(β)|0i.

We have
D† (β) = D−1 (β) = D(−β).

The completeness relation is given by

Z
1
d2 β|βihβ| = I
π C

where I is the identity operator. Let |βi, |γi be coherent states. Then
 
1
hγ|βi = exp − (|β|2 + |γ|2 ) + βγ ∗ .
2

It follows that
∗
hβ| − βi = e−2ββ .

Let |ni be number states. Then

1 2 βn
 
hn|βi = exp − |β| √ .
2 n!

Let n̂ = b† b be the number operator. Then

hβ|n̂|βi = ββ ∗ .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 423

Coherent States 423

19.2 Solved Problems

Problem 1. Bose creation b† and annihilation b operators obey the
Heisenberg algebra [b, b† ] = I, [b, b] = [b† , b† ] = 0 with b|0i = 0, where |0i is
the vacuum state. The coherent states |βi can be obtained by applying the
unitary displacement operator

D(β) := exp(βb† − β ∗ b)

on the vacuum state |0i, i.e.

 ∞
1 2 X βn

† ∗
|βi := D(β)|0i = exp(βb − β b)|0i = exp − |β| √ |ni, β ∈ C.
2 n=0 n!

Show that from this definition the coherent states can also be obtained as
the eigenstates of the annihilation (destruction) operator b, i.e. b|βi = β|βi.

Solution 1. We find for the commutator

[b, (βb† − β ∗ b)n ] = βn(βb† − β ∗ b)n−1 , n = 1, 2, . . .

and therefore we have the commutation relation [b, D(β)] = βD(β). Since
b|0i = 0|0i we have

0|0i = D(β)b|0i = (b − βI)D(β)|0i = (b − βI)|βi

where we used the commutation relation given above.

Problem 2. Let D(β) be the displacement operator. Find the operators

D† (β) and D−1 (β).

Solution 2. We find D† (β) = D−1 (β) = D(−β).

Problem 3. Let |βi and |γi be coherent states.

(i) Calculate hγ|βi. Calculate h0|D(β)|0i ≡ h0|βi.
(ii) Find the probability |hγ|βi|2 .

Solution 3. (i) Since

 ∞  ∞
1 2 X βn 1 2 X γm
 
|βi = exp − |β| √ |ni, |γi = exp − |γ| √ |mi
2 n=0 n! 2 m=0 m!

we find
∞ X ∞
γ ∗m β n
 X
1
hγ|βi = exp − (|β|2 + |γ|2 ) √ √ hm|ni
2 m=0 n=0 m! n!
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 424

424 Problems and Solutions

∞
(βγ ∗ )n
 X
1 2 2
= exp − (|β| + |γ| )
2 n=0
n!
 
1
= exp − (|β|2 + |γ|2 ) + βγ ∗
2
where we used hm|ni = δmn . Using this result we find
 
1
h0|βi = exp − |β|2 .
2

(ii) From (i) we obtain |hγ|βi|2 = exp(−|β − γ|2 ). If γ = β we have

2
|hβ|βi|2 = 1. If γ = −β we obtain |hβ| − βi|2 = e−4|β| .

Problem 4. Let |βi be a coherent state. Let n̂ := b† b. Calculate

hβ|n̂|βi.

Solution 4. Since b|βi = β|βi, hβ|b† = hβ|β ∗ we obtain

hβ|n̂|βi = ββ ∗ = |β|2 .

Problem 5. Let b, b† be Bose annihilation and creation operators and x̂,

p̂ be canonical position and momentum operators which are related by the
equations
1 1
b = √ (x̂ + ip̂), b† = √ (x̂ − ip̂).
2 2
Express the displacement operator with x̂ and p̂.

Solution 5. We obtain

D(x̂, p̂) = exp(i(px̂ − xp̂))

where the real and complex parameters are also related by

1 1
β = √ (x + ip), β̄ = √ (x − ip).
2 2

Problem 6. Consider the Hamilton operator Ĥ = ~ωb† b. Let

U (t) := exp(−itĤ/~).

Find the state U (t)|βi, where |βi is a coherent state.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 425

Coherent States 425

Solution 6. Since
∞
2 X βn
|βi = e−|β| /2
√ |ni
n=0 n!

and b† b|ni = n|ni we find

U (t)|βi = |βe−iωt i.

Thus the linear evolution of |βi is a rotation in phase space. The initial
state will be revived at ωt = 2π, 4π, . . . as expected.

Problem 7. Let D(β) := exp(βb† − β ∗ b) be the displacement operator.

(i) Find the operators D(β)bD(−β), D(β)b† D(−β).
(ii) Find the operators D† (β)bD(β), D† (β)b† D(β).

Solution 7. (i) Since [b† , b] = −I, [b† , [b† , b]] = 0 we obtain

D(β)bD(−β) = b − βI, D(β)b† D(−β) = b† − β ∗ I.

Note that D(−β) = D† (β) = D−1 (β).

(ii) Using parameter differentiation yields

D† (β)bD(β) = b + βI, D† (β)b† D(β) = b† + β ∗ I.

Problem 8. Coherent states are defined as

∞
∗ X βm
|βi := e−ββ /2
√ |mi
m=0 m!

where β is a complex number. We have b|βi = β|βi (eigenvalue equation),

∞
X
|mihm| = I
m=0

(completeness relation) and b† b|ni = n|ni.

(i) Calculate hn|βi and then the probability Pn (β) := |hn|βi|2 .
(ii) Let n̂ := b† b. Calculate hn̂i := hβ|n̂|βi, hn̂2 i := hβ|n̂2 |βi.
(iii) Calculate the variance h(∆n̂)2 i := h(n̂ − hn̂iI)2 i.

Solution 8. (i) Using hn|mi = δnm we have

  n
1 β
hn|βi = exp − |β|2 √ .
2 n!
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 426

426 Problems and Solutions

Thus we obtain
(ββ ∗ )n exp(−ββ ∗ )
Pn (β) = .
n!
This is a Poisson distribution.
(ii) Using b|βi = β|βi and therefore hβ|b† = hβ|β ∗ we find

hn̂i = hβ|b† b|βi = ββ ∗ , hn̂2 i = (ββ ∗ )2 + ββ ∗

where we used bb† = b† b + I.

(iii) Applying the results from (ii) we obtain

h(∆n̂)2 i := h(n̂ − hn̂iI)2 i = ββ ∗ .

Problem 9. Let |0i, |1i, . . . be the number states. An arbitrary normal-

ized state |gi can be expanded as
∞
X ∞
X
|gi = cj |ji, c∗j cj = 1.
j=0 j=0

Express the state using coherent states |βi. Consider the special case that
cj = 0 for all j except for cn = 1.

Solution 9. We find
Z
1 2
|gi = e−|β| /2
g(β ∗ )|βid2 β
π C

where d2 β = d(<(β))d(=(β)) and

∞
X cj (β ∗ )j
g(β ∗ ) = √ .
j=0
j!

If all cj are equal to 0 except for cn = 1 we obtain

Z ∗ n
1 2
/2 (β )
|gi = |ni = e−|β| √ |βid2 β.
π C n!

Problem 10. Let |βi be a coherent state and |ψi be an arbitrary state
in the Hilbert space containing |βi. Show that
 
1 2
|ψ(β)| ≤ exp |β| .
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 427

Coherent States 427

Solution 10. We have the following identity (completeness relation)

Z ∞
1 2
X
d β|βihβ| = |mihm| = I
π C m=0

from which it follows that the system of coherent states is complete. Using
this equation we can expand an arbitrary state |ψi with respect to the state
|βi Z
1
|ψi = d2 βhβ|ψi|βi.
π C
If the coherent state |βi is taken as |ψi, then this equation defines a linear
dependence between the different coherent states. It follows that the system
of coherent states is supercomplete, i.e. it contains subsystems which are
complete. Using the definition for the coherent state given above we obtain
∞
βn
 
1 X
hβ|ψi = exp − |β|2 ψ(β ∗ ), ψ(β) = √ hn|ψi.
2 n=0 n!

The inequality |hn|ψi| ≤ 1 means that the function ψ(β) for the normaliza-
tion state |ψi is an entire analytic function of the complex variables β. We
also have |hβ|ψi| ≤ 1. Therefore we find a bound on the growth of ψ(β)
 
1 2
|ψ(β)| ≤ exp |β| .
2
The normalization condition can now be written as
Z
1
d2 β exp(−|β|2 )|ψ(β)|2 = hψ|ψi.
π C
The expansion of an arbitrary state |ψi with respect to coherent states now
takes the form
Z  
1 1 2
|ψi = d β exp − |β| ψ(β ∗ )|βi.
2
π C 2

Problem 11. Coherent states |βi can be written as |βi = D(β)|0i, where
D(β) is the displacement operator and |0i denotes the vacuum state. Show
that
D(β)D(γ) = exp(i=(βγ ∗ ))D(β + γ). (1)

Solution 11. Since [b, b† ] = I we have

[βb† −β ∗ b, γb† −γ ∗ b] = −[βb† , γ ∗ b]−[β ∗ b, γb† ] = (βγ ∗ −β ∗ γ)I = 2i=(βγ ∗ )I.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 428

428 Problems and Solutions

Using the Baker-Campbell-Hausdorff formula

eA eB = eA+B e[A,B]/2
for [[A, B], A] = 0 and [[A, B], B] = 0, we find (1). As a consequence we
have ∗
D(β)D(γ)D(−β) = e2i=(βγ ) D(γ).
It follows that the operators exp(2πit)D(β) form a group. An element g of
this group is defined by the real number t and a complex number β: g(t, β).
The product of two group elements g = g1 g2 is given by g(t, β) with
1
t = t1 + t2 + =(β2 β1∗ )
2π
and β = β1 + β2 .

Problem 12. Consider the displacement operator D(β). Show that

∗
−β ∗ γ
D(β)D(γ) = eβγ D(γ)D(β). (1)

Solution 12. The Baker-Campbell-Hausdorff formula

eA+B = eA eB e−[A,B]/2 = eB eA e[A,B]/2
for [A, [A, B]] = [B, [A, B]] = 0 can be applied since
[βb† − β ∗ b, γb† − γ ∗ b] = (βγ ∗ − β ∗ γ)I
where I is the identity operator. Thus (1) follows.

Problem 13. Let |βi be a coherent state. Show that

Z
1
|βihβ|d2 β = I
π C
where I is the identity operator and the integration is over the entire com-
plex plane. Set β = r exp(iφ) with 0 ≤ r < ∞ and 0 ≤ φ < 2π.

Solution 13. We have

∞ ∞
1 X X |nihm|
Z Z
1 2
|βihβ|d2 β = √ e−|β| β ∗m β n d2 β.
π C π n=0 m=0 n! m! C

Using β = r exp(iφ) we arrive at

∞ ∞
1 X X |nihm| ∞ −r2 n+m
Z Z Z 2π
1
|βihβ|d2 β = √ re r dr ei(n−m)φ dφ.
π C π n=0 m=0 n! m! 0 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 429

Coherent States 429

Since Z 2π
ei(n−m)φ dφ = 2πδnm
0
we have
∞
|nihn| ∞ −s n
Z Z
1 X
|βihβ|d2 β = e s ds
π C n=0
n! 0

2
where we set s = r and therefore ds = 2rdr. Thus
Z ∞
1 2
X
|βihβ|d β = |nihn| = I
π C n=0

where we used the completeness relation for the number states.

Problem 14. (i) The Husimi distribution of a coherent state γ is given

by
ρH 2
γ (β) := |hβ|γi| .

Calculate ρH
γ (β).
(ii) The Husimi distribution of the number state |ni is given by

ρH 2
|ni (β) := |hβ|ni| .

Calculate ρH|ni (β).

(iii) Consider the state |n1 i ⊗ |n2 i. Find

ρH 2
|n1 i⊗|n2 i (β) = |(hβ1 | ⊗ hβ2 |)(|n1 i ⊗ |n2 i)| .

Solution 14. (i) Since

∞ ∞
2 X βn 2 X γn
|βi = e−|β| /2
√ |ni, |γi = e−|γ| /2
√ |ni
n=0 n! n=0 n!

we find 2
/2 −|γ|2 /2 β ∗ γ
hβ|γi = e−|β| e e
where we used hm|ni = δmn . Thus |hβ|γi|2 = exp(−|β − γ|2 ) and the
Husimi distribution of a coherent state is Gaussian.
(ii) Since hm|ni = δmn we find
2 β ∗n
hβ|ni = e−|β| /2
√
n!
and hence 2
e−|β| (|β|2 )n
2
|hβ|ni| = .
n!
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 430

430 Problems and Solutions

The Husimi distribution represents a Poisson distribution over the photon

number states.
(iii) Since
(hβ1 | ⊗ hβ2 |)(|n1 i ⊗ |n2 i) = hβ1 |n1 ihβ2 |n2 i
we obtain
2 2
e−|β1 | (|β1 |2 )n1 e−|β2 | (|β2 |2 )n2
ρH
|n1 i⊗|n2 i (β) = .
n1 ! n2 !

Problem 15. In a Kerr medium the state evolution is governed by the

interaction Hamilton operator

Ĥ = κ(b† b)2

where κ is a coupling constant proportional to the nonlinear susceptibility

of the medium. A coherent input signal state |βi evolves according to the
solution of the Schrödinger equation

|ψc (t)i = exp(−iĤt)|βi.

Calculate |ψc (t)i for t = π/(2κ). Discuss.

Solution 15. Straightforward calculation yields

1
|ψc (t = π/(2κ))i = √ (e−iπ/4 |βi + eiπ/4 | − βi)
2
1 −iπ/4
= √ (e D(β) + eiπ/4 D(−β))|0i
2

where D(β) is the displacement operator. The state describes a super-

position of two coherent states with opposite phases. When |β| becomes
large the two components become mesoscopically distinguishable states of
the radiation field. Realistic values of the Kerr nonlinear susceptibilities
are quite small, thus requiring a long interaction time, or equivalently a
large interaction length. Thus losses become significant and the resulting
decoherence may destroy the quantum superposition.

Problem 16. Let |βi be a coherent state. Express the density operator
Z 2π
1
ρ= dφ||β|e−iφ ih|β|e−iφ |
2π 0

with number states |ni.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 431

Coherent States 431

Solution 16. We obtain

∞
2 X |β|2n
ρ = e−|β| |nihn|.
n=0
n!

This is a collection of number states.

Problem 17. Let b† , b be Bose creation and annihilation operators and

n̂ = b† b. We can introduce nonlinear coherent states B and B † with an
intensity dependent function f (n̂) (which is an operator valued analytic
function)

B := bf (n̂) = f (n̂ + I)b, B † := f † (n̂)b† = b† f † (n̂ + I).

We consider f to be real and non-negative, i.e. f † (n̂) = f (n̂).

(i) Calculate the commutator [B, B † ]. Then consider the special case
f (n̂) = I and f (n̂) = n̂.
(ii) Express the harmonic oscillator for B † and B
1
Ĥ = (BB † + B † B)
2
using n̂ and f (n̂).
(iii) Find the eigenstate of the operator B, i.e., find the nonlinear coherent
state
B|ziN L = λ|ziN L .
What is the condition that the state belongs to the Fock space?

Solution 17. (i) Since f is analytic we have a Taylor expansion. We

obtain
[B, B † ] = (n̂ + I)f (n̂ + I)f (n̂ + I) − n̂f (n̂)f (n̂).
With f (n̂) = I we obtain the result for the Bose operators b and b† . For
the case f (n̂) = n̂ we obtain

[B, B † ] = 3n̂2 + 3n̂ + I.

(ii) We obtain
1
Ĥ = ((n̂ + I)f (n̂ + I)f (n̂ + I) + n̂f (n̂)f (n̂)).
2
(iii) Using the expansion with respect to the Fock basis { |0i, |1i, . . . } and
the fact that f is an analytic function we obtain
∞
X
|ziN L = Nf (|z|2 )−1/2 Cn z n |ni
n=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 432

432 Problems and Solutions

where the coefficients Cn are given by

1
Cn = p , C0 = 1, [f (n)]! := f (n)f (n − 1) · · · f (1)
[nf (n)f (n)]!

and the normalization constant is given by

∞
X
Nf (|z|2 ) = |Cn |2 |z|2n .
n=0

In order to have states belonging to the Fock space, it is required that

0 < Nf (|z|2 ) < ∞. This implies that

|z| ≤ lim n[f (n)]2 .

n→∞

The function f (n) corresponding to any nonlinear coherent state is found

to be
Cn−1
f (n) = √ .
nCn

Problem 18. Let D(β) be the displacement operator. Is

D(β)D(β 0 ) = D(β + β 0 ) ?

Prove or disprove.

Solution 18. The answer is no. We find

 
0 1
D(β)D(β ) = exp (ββ − β β ) D(β + β 0 ).
0∗ ∗ 0
2

Note that β ∈ C. If β, β 0 ∈ R then we have D(β)D(β 0 ) = D(β + β 0 ).

Problem 19. Consider the coherent state |βi. We define the Schrödinger
cat states

|β+ i := N+ (|βi + |−βi), |β− i := N− (|βi − |−βi).

(i) Normalize the two states.

(ii) Calculate the probabilities |hβ+ |ni|2 , hβ− |ni|2 and discuss.

Solution 19. (i) From hβ+ |β+ i = 1, hβ− |β− i = 1 and

hγ|βi = exp(−|β − γ|2 /2)

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 433

Coherent States 433

we obtain
exp(|β|2 /2) exp(|β|2 /2)
N+ = p , N− = p .
2 cosh(|β|2 ) 2 sinh(|β|2 )

(ii) Since
2 βn
hn|βi = e−|β| /2
√
n!
we obtain the probabilities

1 |β|2n
|hβ+ |ni|2 = 2
(1 + (−1)n )
cosh(|β| ) n!
1 |β|2n
|hβ− |ni|2 = 2
(1 − (−1)n ).
sinh(|β| ) n!

Problem 20. Consider the linear operator

Z := b1 + b†2 ≡ b ⊗ I + I ⊗ b (1)
†
−z ∗ b
where b1 = b ⊗ I and b†2 = I ⊗ b† . Let Db (z) := ezb (z ∈ C) be the
displacement operator and
∞
1 X
|0ii := √ (−1)n |ni ⊗ |ni.
π n=0

The states |zii are defined by

|zii := Db1 (z)|0ii = Db2 (z ∗ )|0ii.

Find the state Z|zii. Discuss.

Solution 20. We have [Z, Z † ] = 0 and

Z|zii = z|zii, z ∈ C.

Thus |zii is an eigenstate of Z. For z = 0 the state |0ii can be approximated

by a physical (normalizable) state called the twin beam state - corresponding
to the output of a non-degenerate optical parametric amplifier in the limit
of infinite gain.

Problem 21. Consider the beam splitter interaction given by the unitary
transformation
UBS = exp(iθ(b1 b†2 + b†1 b2 ))
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 434

434 Problems and Solutions

where b1 and b2 are the Bose annihilation operators. Let |βi, |γi be coherent
states. Calculate
UBS (|γi ⊗ |βi).

Solution 21. We obtain

UBS (|γi ⊗ |βi) = | cos(θ)γ + i sin(θ)βi ⊗ | cos(θ)β + i sin(θ)γi

where cos2 (θ) (sin2 (θ)) is the reflectivity (transmissivity) of the beam split-
ter.

Problem 22. The normalized Schrödinger cat state of a single-mode

radiation field is given by
1
|β, θi = p (|βi + eiθ | − βi).
2 + 2 cos(θ)e−2|β|2

Discuss the special cases θ = 0, θ = π, θ = π/2.

Solution 22. For θ = 0 we obtain the even coherent states

1
|β, 0i = √ (|βi + | − βi).
2 + 2e−2|β|2
For θ = π we obtain the odd states
1
|β, πi = √ (|βi − | − βi).
2 − 2e−2|β|2
For θ = π/2 we obtain the Yurke-Stoler states
1
|β, π/2i = √ (|βi + i| − βi).
2

Problem 23. We know that

†
−β̄b
|βi = D(β)|0i = eβb |0i
† †
= eβb e−β̄b e−β β̄I/2 |0i = e−β β̄/2 eβb |0i
∞
X βn
= e−β β̄/2 √ |ni
n=0 n!

where
(b† )n
|ni = √ |0i.
n!
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 435

Coherent States 435

Let σ1 be the Pauli spin matrix. Extend the calculation to the state
 
1
D(β)(|0i ⊗
e )
0
where
D(β)
e := exp(βb† ⊗ σ1 − β̄b ⊗ σ1 ).

Solution 23. First we note that since σ12 = I2 and [b, b† ] = I

−[βb† ⊗ σ1 , −β̄b ⊗ σ1 ] = −β β̄IB ⊗ I2 .

Thus we obtain
   
1 βb† ⊗σ1 −β̄b⊗σ1 −β β̄IB ⊗I2 /2 1
D(β)|0i
e ⊗ =e e e |0i ⊗
0 0
 
† 1
= e−β β̄/2 eβb ⊗σx e−βb⊗σ1 |0i ⊗
0
 
† 1
= e−β β̄/2 eβb ⊗σ1 |0i ⊗
0
∞
!  
X β 2n 1
−β β̄/2
=e |2ni ⊗
0
p
n=0 (2n)!
∞
! !
β 2n+1

X 0
+ |2n + 1i ⊗ .
1
p
n=0 (2n + 1)!

Problem 24. Bose creation (b† ) and annihilation (b) operators, where

b† = (b†1 , b†2 , . . . , b†N ), b = (b1 , b2 , . . . , bN )

obey the Heisenberg algebra

[bj , b†k ] = δjk I, [bj , bk ] = [b†j , b†k ] = 0, j, k = 1, . . . , N.

Coherent states, where z ∈ Cn , are defined as eigenvectors of the annihila-

tion operators, that is b|zi = z|zi.
(i) Show that the normalized coherent states are given by
 
1 2
|zi = exp − |z| exp(z · b† )|0i (1)
2
where
N N
zj b†j
X X
|z|2 = |zj |2 , z · b† =
j=1 j=1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 436

436 Problems and Solutions

and |0i = |0 0 . . . 0i is the vacuum vector satisfying b|0i = 0|0i.

(ii) Let |wi be a coherent state. Find hz|wi, hz|wi|2 .
(iii) Calculate Z
dµ(z)|zihz|.
R2N

Solution 24. (i) Consider the number representation

(b† )n1 (b†2 )n2 (b† )nN

|ni ≡ |n1 , n2 , . . . , nN i = √1 √ . . . √N |0i.
n1 ! n2 ! nN !
We expand |zi with respect to |ni and apply
√
bj |ni = nj |n1 , · · · , nj − 1, · · · , nN i

b†j |ni = nj + 1|n1 , · · · , nj + 1, · · · , nN i

p

we find, after normalization, that |zi is given by (1).

(ii) We find  
1 2 2 ∗
hz|wi = exp − (|z| + |w| − 2z · w)
2
and
|hz|wi|2 = exp(−|z − w|2 ).
(iii) Since
N
1Y
dµ(z) = d(<zj )d(=zj )
π j=1

we find Z
dµ(z)|zihz| = I
R2N
where I is the identity operator.

Programming Problem

Problem 1. Give a computer algebra implementation that implements

b|βi = β|βi, hβ|b† = hβ|β ∗ .

Then find b2 |βi, hβ|, hβ|βi, b|βi, hβ|b|βi.

Solution 1. In the SymbolicC++ program bd denotes b† , b denotes b,

cs denotes the coherent state |βi and ds denotes hβ| and conj denotes the
complex conjugate and z denotes β.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 437

Coherent States 437

// coherent.cpp
#include <iostream>
#include "symbolicc++.h"
using namespace std;

int main(void)
{
Symbolic b("b"), bd("bd"), cs("cs"), ds("ds");
b = ~b; bd = ~bd; cs = ~cs; ds = ~ds; // noncommutative
Symbolic z("z"), w("w"), conj("conj");
Equations rules =
(b*cs[z]==z*cs[z],b*cs[w]==w*cs[w],
ds[z]*bd==ds[z]*conj[z],ds[w]*bd==ds[w]*conj[w],
ds[z]*cs[z]==1,ds[w]*cs[w]==1,
ds[w]*cs[z]==exp(-(z*conj[z]+w*conj[w]-2*conj[w]*z)/2),
ds[z]*cs[w]==exp(-(z*conj[z]+w*conj[w]-2*conj[z]*w)/2));
// example 1
Symbolic r1 = b*(b*cs[z]);
r1 = r1.subst_all(rules);
cout << r1 << endl;
r1 = r1[z==1];
cout << r1 << endl;
// example 2
cout << (ds[z]*cs[z]).subst_all(rules) << endl;
// example 3
Symbolic r2 = b*cs[z];
Symbolic r3 = ds[w]*r2;
cout << r2.subst_all(rules) << endl;
cout << r3.subst_all(rules) << endl;
return 0;
}

The output is
z^(2)*cs[z]
cs[1]
1
z*cs[z]
e^(-1/2*z*conj[z]-1/2*w*conj[w]+conj[w]*z)*z

19.3 Supplementary Problems

Problem 1. Let |βi (β ∈ C) be a coherent state. Show that the projection

operator |βihβ| can be expressed as

|βihβ| = exp(−ββ) exp(βb† )|0ih0| exp(βb).

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 438

438 Problems and Solutions

Find tr(|βihβ|.

Problem 2. Let D(β) be the displacement operator.

(i) Show that (normally ordered form of D(β))
 
1 2
D(β) = exp − |β| exp(βb† ) exp(−β ∗ b).
2

(ii) Show that (antinormally ordered form of D(β))

 
1 2
D(β) = exp |β| exp(−β ∗ b) exp(βb† ).
2

Problem 3. Let n̂ = b† b and D(β) be the displacement operator. Find

the commutator
[n̂, D(β)].
Find the commutator
[|nihn|, D(β)].
Find the commutator
[|nihn|, |βihβ|].

Problem 4. Show that the solutions of the eigenvalue problem for the
one-dimensional harmonic oscillator
~2 d 2 mω 2 2
 
− + x u(x) = Eu(x)
2m dx2 2

are given by

1  mω 1/4
un (x) = p Hn (x/x0 ) exp(−x2 /(2x20 ))
2n (n!) π~
p
with En = ~ω(n + 1/2), x0 := ~/(mω) and Hn (x/x0 ) are the Hermite
polynomials.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 439

Chapter 20

Squeezed States

20.1 Introduction
Coherent states are not the most general kind of Gaussian wave packet.
They are also not the most general kind of minimum-uncertainty wave
packets, since the minimum uncertainty wave packet satisfies

∆q∆p = ~/2

which only constrains the product of the dispersions ∆q and ∆p, whereas
for coherent states we have that

(∆q)2 = ~/(2ω) and (∆p)2 = ~ω/2.

For squeezed states one does not have this restriction. Unlike a coherent
state, an initial squeezed state does not remain a minimum-uncertainty
state in the course of time under the harmonic oscillator evolution. The
product ∆q∆p oscillates at twice the harmonic oscillator frequency between
a maximum value and a minimum value. Squeezed states possess the prop-
erty that one quadrature phase has reduced fluctuations compared to the
ordinary vacuum. Squeezed states of the electromagnetic field are gener-
ated by degenerate parametric down conversion in an optical cavity. The
ideal squeezed state is defined as

|ζi := S(ζ)|0i

439
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 440

440 Problems and Solutions

where  
1 ∗ 2 1 †2
S(ζ) := exp ζ b − ζb
2 2
is the one-mode squeezing operator with ζ ∈ C. If we set

ζ = seiθ

(s ≥ 0) then s is called the squeezing parameter.

The operator S(ζ) is unitary, since

ζ † 2 ζ∗ 2
 
†
S (ζ) = exp (b ) − b
2 2

and therefore S † (ζ)S(ζ) = I. In disentangled form S(ζ) is given by

   
1 † 2 iθ 1 † †
S(ζ) = exp − (b ) e tanh(s) exp − (b b + bb ) ln(cosh(s))
2  2
1 2 −iθ
× exp b e tanh(s) .
2

Using this result the squeezed states can be expressed with number states
∞ p  n
p X ((2n)!) 1
|ζi = (sech(s)) − exp(iθ) tanh(s) |2ni
n=0
n! 2

where n = 0, 1, 2, . . . and we used that (b† b + bb† )|0i = |0i. The expansion
over the number state basis only contains even components. The state |ζi
is normalized. We also have the eigenvalue equation

(b cosh(s) + b† eiφ sinh(s))|ζi = 0|ζi

and similarly hζ| is the zero-eigenvalue left eigenstate of

b† cosh(s) + be−iθ sinh(s).

The overlap of two squeezed states is given by

1
hξ|ζi = p
cosh(s1 ) cosh(s2 )(1 − e −i(θ 2 −θ1 ) tanh(s ) tanh(s ))
1 2

where ξ = s1 eiθ1 , ζ = s2 eiθ2 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 441

Squeezed States 441

20.2 Solved Problems

Problem 1. Let ζ ∈ C. The squeezing operator is defined by
ζ † 2 ζ∗ 2
 
S(ζ) := exp − (b ) + b
2 2
where ζ ∈ C. We set ζ = seiθ , where s ≥ 0 and θ ∈ R. The single mode
squeezed states |ζi are defined by
|ζi := S(ζ)|0i.
(i) Show that S(ζ) is unitary.
(ii) Find S(ζ) in the disentangled form.
(iii) Calculate the state |ζi = S(ζ)|0i using the number states.

Solution 1. (i) We obtain

ζ † 2 ζ∗ 2
 
S † (ζ) = exp (b ) − b = S(−ζ).
2 2
Thus we have S † (ζ)S(ζ) = I.
(ii) We obtain
   
1 † 2 iθ 1 † †
S(ζ) = exp − (b ) e tanh(s) exp − (b b + bb ) ln(cosh(s))
2 2
 
1 2 −iθ
× exp b e tanh(s) .
2
(iii) Using the result from (ii) we obtain
∞ p  n
p X ((2n)!) 1
|ζi = (sech(s)) − exp(iθ) tanh(r) |2ni
n=0
n! 2

where we used that (b† b + bb† )|0i = |0i.

Problem 2. Consider the one-mode squeezing operator

∗ 2
b −ζ(b† )2 )
S(ζ) = e1/2(ζ
where ζ ∈ C. Calculate
eb = S † (ζ)bS(ζ), eb† = S † (ζ)b† S(ζ).

Solution 2. We obtain
∗
eb = cosh(|ζ|)b − ζ sinh(|ζ|)b† , eb† = cosh(|ζ|)b† − ζ sinh(|ζ|)b.
|ζ| |ζ|
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 442

442 Problems and Solutions

Problem 3. Consider the single-mode squeezing operator

 
1 † 2 1 ∗ 2
S(ζ) = exp ζ(b ) − ζ b .
2 2
Let |0i be the vacuum state. Find the state S(ζ)|0i ≡ |ζi using number
states |ni. Use this result to find the expectation value hζ|b† b|ζi.

Solution 3. Expansion over the number state basis contains only even
components, i.e.
∞  k p
1 X ν (2k)!
S(ζ)|0i ≡ |ζi = √ |2ki
µ 2µ k!
k=0
iθ iθ
where ζ = se , µ = cosh(s), ν = e sinh(s). Using this result and
b† b|2ki = 2k|2ki
we obtain
hζ|b† b|ζi = |ν|2 .

Problem 4. Let S(ζ) be the one-mode squeezing operator (ζ ∈ C)

 
1 ∗ 2 1 † 2
S(ζ) := exp ζ b − ζ(b ) .
2 2
Calculate
S(ζ)bS † (ζ)
   
b †
S(χ) S (ζ) ≡
b† S(ζ)b† S † (ζ)
and show that we can write
S(ζ)bS † (ζ)
   
∗ b
= T (ζ, ζ ) †
S(ζ)b† S † (ζ) b
where the 2 × 2 matrix T (ζ, ζ ∗ ) depends only on ζ and ζ ∗ .

Solution 4. We have S † (ζ) = S(−ζ). Let  ∈ R. We set

∗ 2
1
b −ζ(b† )2 ) 1 ∗ 2
b −ζ(b† )2 )
f1 () = e 2 (ζ be− 2 (ζ
and ∗ 2
1
b −ζ(b† )2 ) † − 12 (ζ ∗ b2 −ζ(b† )2 )
f2 () = e 2 (ζ b e
with the initial conditions f1 ( = 0) = b and f2 ( = 0) = b† . We find the
system of differential equations for f1 and f2 . Using bb† = I + b† b we have
df1 1 1 ∗ 2 † 2 1 ∗ 2 † 2
= − ζe 2 (ζ b −ζ(b ) ) ((b† )2 b − b(b† )2 )e− 2 (ζ b −ζ(b ) )
d 2
= ζf2 ().
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 443

Squeezed States 443

Analogously, we obtain
df2
= ζ ∗ f1 ().
d
In matrix notation we have the system
    
df1 /d 0 ζ f1
= .
df2 /d ζ∗ 0 f2

Let  
0 ζ
A= .
ζ∗ 0
We find
   
cosh(s) ζ sinh(s)/s cosh(s) eiθ sinh(s)
exp(A) = =
ζ ∗ sinh(s)/s cosh(s) e −iθ
sinh(s) cosh(s)

where we used ζ = s exp(iθ). Taking into account the initial conditions we

obtain
S(ζ)bS † (ζ)
    
cosh(s) eiθ sinh(s) b
= .
S(ζ)b† S † (ζ) e−iθ sinh(s) cosh(s) b†

The matrix on the right-hand side can be decomposed as

 
cosh(s) eiθ sinh(s)
≡
e−iθ sinh(s) cosh(s)
   
0 eiθ/2 cosh(s) sinh(s) 0 eiθ/2
.
e−iθ/2 0 sinh(s) cosh(s) e−iθ/2 0

Problem 5. For generating a squeezed state of one mode we start from

the Hamilton operator

Ĥ = ~ωb† b + i~Λ(b2 e2iωt − (b† )2 e−2iωt ).

A photon of the driven mode, with frequency 2ω, splits into two photons of
the mode of interest, each with frequency ω. Solve the Heisenberg equation
of motion for b and b† .

Solution 5. The Heisenberg equation of motion yields

db 1
= [b, Ĥ](t) = −iωb(t) − 2Λb† (t)e−2iωt
dt i~
db† 1
= [b† , Ĥ](t) = iωb† − 2Λbe2iωt .
dt i~
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 444

444 Problems and Solutions

Using b = eb exp(−iωt) we obtain

deb deb†
= −2Λeb† (t), = −2Λeb(t).
dt dt
Introducing the operators

ebP := 1 (eb + eb† ), ebQ := 1 (eb − eb† )

2 2i
i.e.
eb = ebP + iebQ , eb† = ebP − iebQ

we finally arrive at the system of differential equations

dbP dbQ
= −2ΛbP (t), = 2ΛbQ (t)
dt dt
with the solution of the initial value problem

bP (t) = bP (0)e−2Λt , bQ (t) = bQ (0)e2Λt .

If the state is the vacuum state |0i we obtain a squeezed state. For the
electric field E(r, t) we have

E(r, t) = iE(b(t)eik·r − b† (t)e−ik·r )

= iE(eb(t)eik·r−iωt − eb† (t)e−ik·r−iωt )
= −2E(bP (t) sin(k · r − ωt) + bQ (t) cos(k · r − ωt)).

Thus bP and bQ are the amplitudes of two quadrature components of the

electric field. They are measurable by phase sensitive detection.

Problem 6. Consider the squeezing operator

 
1 2
S(s) := exp s(b − b†2 )
2
where s ∈ R. Find the operators

S(s)q̂S † (s), S(s)p̂S † (s)

where
1 i
q̂ := √ (b + b† ), p̂ := − √ (b − b† ).
2 2

Solution 6. We obtain

S(s)q̂S † (s) = e−s q̂, S(s)p̂S † (s) = es p̂.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 445

Squeezed States 445

Problem 7. Let ζ ∈ C and

1 † †
Ĝ := (ζb b − ζ ∗ bb).
2

(i) Calculate the commutators [Ĝ, b], [Ĝ, [Ĝ, b]].

(ii) Let  
1 † †
S(ζ) := exp (ζb b − ζ ∗ bb) ≡ exp Ĝ.
2
Find S(ζ)bS(−ζ), S(ζ)b† S(−ζ).

Solution 7. (i) We have

1 † †
[Ĝ, b] = ζ[b b , b] = −ζb†
2
and
[Ĝ, [Ĝ, b]] = [Ĝ, −ζb† ] = −ζ[Ĝ, b† ] = |ζ|2 b.
(ii) Using the results from (i), we find

S(ζ)bS(−ζ) = (cosh(s))b − eiθ (sinh(s))b†

where ζ = reiφ . Likewise, we find

S(ζ)b† S(−ζ) = (cosh(s))b† − e−iθ (sinh(s))b.

Problem 8. Consider the unbounded linear operators

1 †2 1 2 1 † 1
K+ := b , K− := b , K0 := (b b + I)
2 2 2 2

A† := b† , A := b
where I is the identity operator.
(i) Show that these operators form a Lie algebra.
(ii) Consider the operator

P := ζK+ − ζ ∗ K− + αA† − α∗ A

where ζ and α are complex numbers. Let

†
V = eβK+ eA eγK0 eνI eδK− eηA

where β, , γ, ν, δ and η are complex numbers. Let eP = V . Find γ, β, δ,

, η, ν as functions of ζ and α.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 446

446 Problems and Solutions

Solution 8. (i) For the commutators we obtain

[K0 , K± ] = ±K± , [K+ , K− ] = −2K0 , [K+ , A] = −A† , [K− , A† ] = A

1 † 1
[K0 , A† ] =
A , [K0 , A] = − A, [A, A† ] = I.
2 2
This Lie algebra refers to squeezed coherent states.
(ii) We write the complex numbers ζ and α in terms of real numbers λ, µ,
θ and φ as ζ = reiθ , α = seiφ . We use the formula
∞
X [A, B]n
eA Be−A =
n=0
n!

where the repeated commutator [A, B]n is defined by

[A, B]n := [A, [A, B]n−1 ]

with [A, B]0 := B. We find

s
eP Ae−P = cosh(r)A − eiθ sinh(r)A† + ((cosh(r) − 1)ei(θ−φ) − sinh(r)eiφ )
r
s
eP A† e−P = cosh(r)A† −e−iθ sinh(r)A+ ((cosh(r)−1)ei(φ−θ) −sinh(r)e−iφ ).
r
The corresponding similarity transformations, induced by the operator V ,
are

V AV −1 = e−γ/2 (A − βA† − I)

V A† V −1 = (eγ/2 − βδe−γ/2 )A† + δe−γ/2 A + ηI − δe−γ/2 I.

From
eP Ae−P = V AV −1 , eP A† e−P = V A† V −1
we find, by separating out terms with A† , A and I, that

γ = −2 ln(cosh(r))
β = eiθ tanh(r)
δ = −e−iθ tanh(r)
s
=− ((cosh(r) − 1)ei(θ−φ) − sinh(r)eiφ )
r cosh(r)
s
η=− ((cosh(r) − 1)e−i(θ−φ) − sinh(r)e−iφ ).
r cosh(r)
The coefficient ν cannot be found by this method. How can we determine
ν? One finds
s2
ν=− ((cosh(r) − 1) + i sin(θ − 2φ)(sinh(r) − r cosh(r))).
r2 cosh(r)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 447

Squeezed States 447

Problem 9. (i) Show that the operators

 
1 1 1 1
K+ = b† b† , K− = bb, K0 = †
b b+ I
2 2 2 2
are generators of the Lie algebra su(1, 1).
(ii) Let D(β) be the displacement operator and

S(ζ) = exp(ζK+ − ζ ∗ K− ), ζ ∈ C.

We define the state

|(β, ζ)i := D(β)S(ζ)|0i.
Show that the ordering D(β)S(ζ) versus S(ζ)D(β) is unitarily equivalent.

Solution 9. (i) We have [K0 , K+ ] = K+ , [K0 , K− ] = −K− , [K+ , K− ] =

−2K0 .
(ii) We have D(β)S(ζ) = S(ζ)D(γ), γ = β cosh(s) − β ∗ eiθ sinh(s) and
z = seiθ .

Problem 10. Consider the squeeze operator

 
1 ∗ 2 †2
S(ζ) = exp (ζ b − ζb )
2
with ζ = |ζ| exp(i(φ + θ)). Is S(ζ 0 )S(ζ 00 ) = S(ζ 0 + ζ 00 )?

Solution 10. The answer is no. We find

  
0 00 † 1
S(ζ )S(ζ ) = exp −iφ b b + I S(ζ) (1)
2
with

ζ 0 = |ζ 0 | exp(iθ0 ), µ0 = cosh |ζ 0 |, ν 0 = exp(iθ0 ) sinh |ζ 0 |

and |µ0 |2 − |ν 0 |2 = 1 with analogous relations for ζ 00 , µ00 and ν 00 . In (1) ζ is

obtained from
 
−1 |ν|
ζ = |ζ| exp(i(φ + θ)), |ζ| = tanh
|µ|

with µ = µ0 µ00 + ν 0∗ ν 00 , ν = µ0∗ ν 00 + ν 0 µ00 and |µ|2 − |ν|2 = 1.

Problem 11. (i) A single mode squeeze operator is defined by

 
1
S(z) = exp (z(b† )2 − z ∗ b2 )
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 448

448 Problems and Solutions

where z ∈ C with z = reiθ . Write S(z) in the form S(z) = exp(iH(z)).

Find H(z).
(ii) Calculate
R† (φ)H(z)R(φ)
where R(φ) = exp(iφb† b) is a single-mode rotation. Set φ = θ/2 and thus
find H(r).
(iii) Use the transformation
x + ip x − ip
b= √ , b† = √
2 2
and express H(r) in terms of x and p.
(iv) Discuss the H(r) as a Hamilton function.
(v) Write down the eigenvalue equation Hψ(x) = Eψ with p = −id/dx.
Discuss the spectrum.

Solution 11. (i) Obviously we have

1
H(z) = (z(b† )2 − z ∗ b2 ).
2i
(ii) We find
R† (φ)H(z)R(φ) = H(ze−2iφ ).
Thus for φ = θ/2 we obtain H(r).
(iii) Inserting the transformation yields
r
H(r) = − (xp + px).
2
(iv) The Hamilton function H(r) = −rxp leads to the Hamilton equations
of motion
dx dp
= −rx, = rp.
dt dt
This describes damping of x and the pumping of p. This is a classical
picture of a squeezing process.
(v) Since
r dψ r
H(r)ψ(x) = − (xp + px)ψ(x) = irx + i ψ(x)
2 dx 2
it follows that the eigenvalue equation is
 
d E 1
x ψ(x) = − i + ψ(x).
dx r 2

H(r) has a purely continuous spectrum covering the whole real axis. H(r)
is also parity invariant (x → −x). Therefore each generalized eigenvalue
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 449

Squeezed States 449

E ∈ R is doubly degenerated. The eigenvectors in the sense of generalized

functions are
E 1 −(iE/r+1/2)
ψ± (x) = √ x±
2πr
where  λ 
x for x ≥ 0 0 for x ≥ 0
xλ+ := , xλ− :=
0 for x < 0 |x|λ for x < 0
and λ ∈ C.

Problem 12. Consider the linear operator

1
D̂ := (q̂ p̂ + p̂q̂).
2
We set ~ = 1.
(i) Find the commutators [D̂, q̂] and [D̂, p̂].
(ii) We consider the linear operator S := exp(−iD̂),  ∈ R. We define
∞
X (i)n
S† q̂S := exp(iadD̂)q̂ ≡ (adD̂)n q̂ (1)
n=0
n!

where
(adD̂)q̂ := [D̂, q̂].
Calculate the operators and S† p̂S .
S† q̂S
(iii) Let
√   √  
mω p̂ † mω p̂
b= √ q̂ + i , b = √ q̂ − i .
2 mω 2 mω

Express D̂ in terms of b and b† .

(iv) Consider the normalized state |i := S |0i. Calculate the expectation
values
h|q̂|i, h|p̂|i, h|q̂ 2 |i, h|p̂2 |i.

Solution 12. (i) Since [q̂, p̂] = iI we find

[D̂, q̂] = −iq̂, [D̂, p̂] = ip̂.

(ii) Using the result from (i) and the definition (1) we find the operator

S† q̂S = e q̂.

Using the result from (i) and the definition (1) with q̂ replaced by p̂, we
find
S† p̂S = e− p̂.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 450

450 Problems and Solutions

(iii) First we express q̂ and p̂ in terms of b and b† . Using the commutation

relation [q̂, p̂] = iI, we obtain

D̂ = 2i(b† b† − bb).

(iv) Using the results from (i) through (iii) we find the expectation values

h|q̂|i = h0|S† q̂S |0i = e h0|q̂|0i = 0

h|p̂|i = h0|S† p̂S |0i = e− h0|p̂|0i = 0

1
h|q̂ 2 |i = e2 h0|q̂ 2 |0i = e2= (∆q)2
2mω
mω
h|p̂2 |i = e−2 h0|p̂2 |0i = e−2 = (∆p)2 .
2

Problem 13. Consider the uncertainty relation

1
(hψ|Â2 |ψi − hψ|Â|ψi2 )(hψ|B̂ 2 |ψi − hψ|B̂|ψi2 ) ≥ |hψ|[Â, B̂]|ψi|2
4

where  and B̂ are observable, [Â, B̂] denotes the commutator and |ψi is
a normalized state. Let b† , b be Bose creation and annihilation operators
and
1 1
 = √ (ib − ib† ), B̂ = √ (b + b† ).
2 2
(i) Let |ψi = |βi be a coherent state (β ∈ C). Find the left-hand and
right-hand side of the uncertainty relation.
(ii) Let |ψi = |ni be a number state (n = 0, 1, 2, . . .). Find the left-hand
and right-hand side of the uncertainty relation.
(iii) Let |ψi = |ζi be a squeezed state (ζ ∈ C). Find the left-hand and
right-hand side of the uncertainty relation.

Solution 13. (i) For the commutator we find

[Â, B̂] = iI

where I is the identity operators. Thus with hβ|βi = 1 we obtain for the
right-hand side
1 1
|hβ|iI|βi|2 = .
4 4
Now b|βi = β|βi, hβ|b† = hβ|β. It follows that

1 2
hψ|Â|ψi2 = (−β 2 − β + 2ββ)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 451

Squeezed States 451

1 2 2
hψ|B̂|ψi2 = (β + β + 2ββ)
2
1 2
hψ|Â2 |ψi = (1 − β 2 − β + 2ββ)
2
1 2
hψ|B̂ 2 |ψi = (1 + β 2 + β + 2ββ)
2
and
1 1
hβ|Â2 |βi − hβ|Â|βi2 = , hβ|B̂ 2 |βi − hβ|B̂|βi2 = .
2 2
It follows that the uncertainty relation is an equality for the present case.
(ii) With hn|ni = 1 for the right-hand side we have again 14 . From

hn|Â|ni = 0, hn|B̂|ni = 0

and
1 1
hn|Â2 |ni =
(1 + 2n), hn|B̂ 2 |ni = (1 + 2n).
2 2
Thus we have the inequality
1 1
(1 + 2n)2 ≥ .
4 4
So if n = 0 we have an equality.
(iii) We set ζ = seiθ . From |ζi = S(ζ)|0i, where S(ζ) is the squeezing
operator we obtain

hζ|b|ζi = h0|S −1 (ζ)bS(ζ)|0i = h0|(cosh(s)b − eiθ sinh(s)b† )|0i = 0

hζ|b† |ζi = h0|S −1 (ζ)b† S(ζ)|0i = h0|(cosh(s)b† − e−iθ sinh(s)b)|0i = 0.

Now

hζ|b2 |ζi = h0|S −1 (ζ)bS(ζ)S −1 (ζ)bS(ζ)|0i = −eiθ cosh(s) sinh(s)

hζ|(b† )2 |ζi = h0|S −1 (ζ)b† S(ζ)S −1 (ζ)b† S(ζ)|0i = −e−iθ cosh(s) sinh(s).
Hence
1 1
hζ|Â2 |ζi = cosh(2s) + sinh(2s) cos(θ)
2 2
and
1 1
hζ|B̂ 2 |ζi = cosh(2s) − sinh(2s) cos(θ)
2 2
and the inequality follows
1 1
(cosh2 (2s) − sinh2 (2s) cos2 (θ)) ≥ .
4 4
With cosh2 (2s) − sinh2 (2s) = 1 we have an equality for θ = 0 and θ = π.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 452

452 Problems and Solutions

Problem 14. Consider the operator

† †
U (z) := ezb1 b2 −z̄b2 b1

where b†1 , b†2 are Bose creation operators and b1 , b2 are Bose annihilation
operators and z ∈ C. Find the operators

U (z)b1 U (z)−1 , U (z)b2 U (z)−1 .

Solution 14. We have

† †
U −1 (z) = e−zb1 b2 +z̄b2 b1 .

Thus
z sin(|z|)
U (z)b1 U (z)−1 = cos(|z|)b1 − b2
|z|
z̄ sin(|z|)
U (z)b2 U (z)−1 = cos(|z|)b2 + b1 .
|z|
We can write
z̄ sin(|z|)
!
−1 −1
cos(|z|) |z|
(U (z)b1 U (z) , U (z)b2 U (z) ) = (b1 , b2 )
− z sin(|z|)
|z| cos(|z|)

where the matrix on the right-hand side is an element of the Lie group
SU (2).

Problem 15. Consider the operator

† †
U (z) := ezb1 b2 −z̄b2 b1

where b†1 , b†2 are Bose creation operators and b1 , b2 are Bose annihilation
operators and z ∈ C. Find the operators

U (z)b1 U (z)−1 , U (z)b†2 U (z)−1 .

Solution 15. We have

† †
U −1 (z) = e−zb1 b2 +z̄b2 b1 .

Thus
z sinh(|z|) †
U (z)b1 U (z)−1 = cosh(|z|)b1 − b2
|z|
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 453

Squeezed States 453

z̄ sinh(|z|)
U (z)b†2 U (z)−1 = cosh(|z|)b†2 − b1 .
|z|
We can write in matrix form

− z̄ sinh(|z|)
!
cosh(|z|)
(U (z)b1 U (z)−1 , U (z)b†2 U (z)−1 ) = (b1 , b†2 ) |z|
− z sinh(|z|)
|z| cosh(|z|)

where the matrix on the right-hand side is an element of the Lie group
SU (1, 1).

Problem 16. Describe how two-mode squeezed state can be generated.

Solution 16. Two mode squeezed states can be generated either by entan-
gling two independent single-mode squeezed states via a 50:50 beam splitter
or by employing the non-degenerate operation of a nonlinear medium in the
presence of two incoming modes. The unitary operator describing two-mode
squeezing is
U12 (ζ) = exp(−i(ζb1 b2 + ζ ∗ b†1 b†2 )/2)
where ζ ∈ C is the squeezing parameter.

Problem 17. Consider the two-mode squeezing operator

S2 (ζ) = exp(ζb†1 b†2 − ζ ∗ b1 b2 ).

Let |0i ⊗ |0i be the two-mode vacuum state. Find the normalized state

S2 (ζ)(|0i ⊗ |0i)

expressed in number states.

Solution 17. We obtain

∞  k
1 X ν
S2 (ζ)(|0i ⊗ |0i) = √ |ki ⊗ |ki
µ µ
k=0

where
ζ = seiθ , µ = cosh(s), ν = eiθ sinh(s).
This state is known as two mode squeezed vacuum or twin beam state.

Problem 18. Let |0i be the vacuum state. Then we define the coherent
squeezed state as
|β, ζi := D(β)S(ζ)|0i
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 454

454 Problems and Solutions

where D(β) is the displacement operator and S(ζ) is the squeezing operator
with β, ζ ∈ C.
(i) Show that the state |β, ζi is normalized.
(ii) Show that S(ζ)D(β)|0i = 6 |β, ζi.

Solution 18. (i) We have

hβ, ζ|β, ζi = h0|S(−ζ)D(−β)D(β)S(ζ)|0i = 1.

(ii) We have

S(ζ)D(α)|0i = S(ζ)D(β)S(−ζ)S(ζ)|0i
= exp(β(b† cosh(s) + be−iφ sinh(s))
−β ∗ (b cosh(s) + b† eiθ sinh(s)))S(ζ)|0i
= D(β cosh(s) − β ∗ eiθ sinh(s))S(ζ)|0i
= |β cosh(s) − β ∗ eiθ sinh(s), ζi

where we used that

S(ζ)D(β)S(−ζ) =
exp(β(b† cosh(s) + be−iθ sinh(s)) − β ∗ (b cosh(s) + b† eiθ sinh(s))).
Thus |β, ζi can be written as

|β, ζi = S(ζ)D(β cosh(s) + β ∗ sinh(s))|0i.

Problem 19. Let n̂ = b† b. Find hβ, ζ|n̂|β, ζi.

Solution 19. We have

hβ, ζ|n̂|β, ζi = h0|(b† cosh(s) − be−θ sinh(s) + β ∗ I)

×(b cosh(s) − b† eiθ sinh(s) + βI)|0i
= sinh2 (s) + |β|2 .

Problem 20. Let |αi be a coherent state. The Husimi distribution of a

quantum wave function |ψi is given by

ρH 2
|ψi (α) = |hψ|αi| .

(i) A coherent squeezed state |γ, βi is defined as

|ζ, βi := D(β)S(ζ)|0i
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 455

Squeezed States 455

where D(β) is the displacement operator and S(ζ) := exp((ζ ∗ b2 −ζ(b† )2 )/2).
The coherent squeezed states |ζ, βi also minimize the uncertainty relation,
however, the variance of both canonically coupled variables are not equal.
The modulus g of the complex number ζ = ge2iθ determines the strength
of squeezing, s = eg − 1, while the angle θ orients the squeezing axis. Find
the Husimi distribution

ρH
|ζ,βi (α) = |hα|ζ, βi|
2

of a coherent squeezed state |ζ, βi.

Solution 20. Let α = α1 + iα2 , where α1 , α2 ∈ R. We obtain

ρH 2 2 2 2 2
|ζ,βi (α) = |hα|ζ, βi| = exp(−(<(β)−α1 ) /(s+1) −(=(β)−α2 ) (s+1) ).

Problem 21. Let (r ∈ R)

 
1 2
S(r) = exp r(b − b†2 ) , D(β) = exp(βb† − β ∗ b).
2

Find the operator S † (r)D(β)S(r).

Solution 21. We find

S † (r)D(β)S(r) = D(<(β)er + i=(β)e−r ).

Problem 22. Let D(β) be the displacement operator

D(β) = exp(βb† − β ∗ b)

and
S(r, φ) = exp(reiφ (b† )2 /2 − re−iφ b2 /2)
be the squeeze operator with the squeeze factor r ≥ 0 and squeeze angle
φ ∈ (−π, π]. Let
∞  n
1 X n
ρT := |nihn|
n + 1 n=0 n + 1

be the Bose-Einstein density operator with the mean occupancy

   −1
~ω
n= exp −1 .
kB T
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 456

456 Problems and Solutions

Consider the displaced squeezed thermal state

ρ = D(β)S(r, φ)ρT S † (r, φ)D† (β). (1)

The Weyl expansion of the density operator

Z
1
ρ= d2 λ χ(λ)D(−λ)
π C

with d2 λ = d<(λ)d=(λ) provides the one-to-one correspondence between

the density operator ρ and its characteristic function
∞
X
χ(λ) := tr(ρD(λ)) ≡ hn|ρD(λ)|ni. (2)
n=0

A Gaussian state has a characteristic function of the form

   
1 1 1
χ(λ) = exp − A + |λ|2 − B ∗ λ2 − B(λ∗ )2 + C ∗ λ − Cλ∗
2 2 2

with A > 0. Find the coefficients A, B, C for the given ρ.

Solution 22. Calculating (2) where ρ is given by (1) and comparing

coefficients yields
   
1 1 1 iφ
A= n+ cosh(2r) − , B = − n + e sinh(2r), C = β.
2 2 2

Problem 23. Let S(ζ) be the one-mode squeeze operator with ζ = |ζ|eiθ .
Let D(β) be the displacement operator. Find S(ζ)D(β)S −1 (ζ).

Solution 23. We obtain

S(ζ)D(β)S −1 (ζ) = D(β),

e βe = cosh(|ζ|)β + eiθ sinh(|ζ|)β ∗ .

Programming Problem

Problem 1. Give a computer algebra implementation of squeezed states.

Solution 1. The Bose creation and annihilation operators are denoted

by b and bd and b*bd==1+bd*b is implemented.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 457

Squeezed States 457

// squeezed.cpp
#include <iostream>
#include "symbolicc++.h"
using namespace std;

int main(void)
{
Symbolic eps("eps"), zeta("zeta"), zetab("zetab");
Symbolic b("b"), bd("bd");
b = ~b; bd = ~bd; // b and bd are noncommutative
Symbolic S, Sd;
Symbolic arg("arg");
arg = -zeta*bd*bd/2 + zetab*b*b/2;
S = exp(eps*arg);
Sd = exp(-eps*arg);
Symbolic result1("result1");
result1 = S*arg*b*Sd - S*b*arg*Sd;
result1 = result1.subst_all(b*bd==1+bd*b);
cout << result1 << endl << endl;
Symbolic result2("result2");
result2 = S*arg*bd*Sd - S*bd*arg*Sd;
result2 = result2.subst_all(b*bd==1+bd*b);
cout << result2 << endl;
return 0;
}

The output is

e^(-1/2*eps*zeta*bd^(2)+1/2*eps*zetab*b^(2))*bd*zeta*
e^(1/2*eps*zeta*bd^(2)-1/2*eps*zetab*b^(2))

e^(-1/2*eps*zeta*bd^(2)+1/2*eps*zetab*b^(2))*b*zetab*
e^(1/2*eps*zeta*bd^(2)-1/2*eps*zetab*b^(2))

20.3 Supplementary Problems

Problem 1. Let |βi be a coherent state and |ζi be a squeezed state.

Calculate
hβ|ζi, |hβ|ζi|2 .

Problem 2. Let n̂ = b† b be the number operator and S(ζ) be the squeezed

operator. Find the commutator

[n̂, S(ζ)].
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 458

458 Problems and Solutions

Problem 3. Consider the displacement operator, squeezing operator and

rotation operator
† † 2 †
−βb ) −ζb2 )/2
D(β) = eβb , S(ζ) = e(ζ(b , R(φ) = e−iφb b

where β, ζ ∈ R. Find the commutators

[D(β), S(ζ)], [D(β), R(φ)], [R(φ), S(ζ)].

Problem 4. Let b† , b be Bose creation and annihilation operators and

γ ∈ R. Calculate
† 2 † 2 † 2
) −b2 ) ) −b2 ) ) −b2 ) † −γ((b† )2 −b2 )
eγ((b be−γ((b , eγ((b b e

utilizing
1 1
eA Be−A = B + [A, B] + [A, [A, B]] + [A, [A, [A, B]]] + · · ·
2! 3!

Problem 5. Let ζ ∈ C. The squeezing operator is defined as

 
1 † 2 1 2
S(ζ) = exp ζ(b ) − ζb .
2 2

We set ζ = seiθ . Show that

S † (ζ)bS(ζ) = cosh(s)b + eiθ sinh(s)b† .

S † (ζ)b† S(ζ) = cosh(s)b† + e−iθ sinh(s)b

hζ|b† b|ζi = h0|S † (ζ)b† S(ζ)S † (ζ)bS(ζ)|0i = sinh2 (s).

Problem 6. Let D(β) be the displacement operator and S(ζ) be the

squeezing operator with

β = reiφ , ζ = seiθ .

Show that

S(ζ)D(β)|0i = D(β cosh(s) − βeiθ sinh(s))S(ζ)|0i.

Problem 7. Let S(ζ), D(β), R(α) be the squeezing operator, displace-

ment operator and phase shift operator, where

R(α) = exp(iαb† b).

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 459

Squeezed States 459

Let ρ be a density operator. Then

ρe = D(β)R(α)S(ζ)ρS † (ζ)R† (α)D† (β)

is a density operator. Calculate ρe for

ρ = |nihn|, ρ = |βihβ|, ρ = |ζihζ|.

Problem 8. Consider the two-mode squeezing operator

S(ζ) = exp(ζb†1 b†2 − ζb1 b2 ), ζ = sei(θ+π/2)

where
b†1 = b† ⊗ I, b†2 = I ⊗ b† .
Show that
∞
1 X
S(ζ)(|0i ⊗ |0i) = ein(θ+π/2 (tanh(s))n |ni ⊗ |ni.
cosh(s) n=0

The state is mode-entangled.

Problem 9. Let b†1 , b†2 , b†3 , b1 , b2 , b3 be Bose creation and annihilation

operators. Note that

[b†1 b2 − b1 b†2 , b†2 b3 − b2 b†3 ] = b†1 b3 − b1 b†3

[b†1 b3 − b1 b†3 , b†1 b2 − b1 b†2 ] = b†2 b3 − b†3 b2 .

Let θ1 , θ2 , θ3 ∈ R and consider the unitary operator

† † † † † †
U (θ1 , θ2 , θ3 ) = eθ3 (b2 b3 −b2 b3 ) eθ2 (b2 b1 −b2 b1 ) eθ1 (b2 b3 −b2 b3 ) .

Find the 3 × 3 matrix M such that

 †  † †  †
b1 U b1 U b1
U (θ1 , θ2 , θ3 )  b†2  =  U b†2 U †  = M (θ1 , θ2 , θ3 )  b†2  .
b†3 U b†3 U † b†3

Problem 10. Consider the unitary operator for two-mode squeezing

U (ζ) = exp(ζb†1 b†2 − ζb1 b2 ).

The Hamilton operator with α ∈ R is given by

Ĥ = iα(b†1 b†2 − b1 b2 ).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 460

460 Problems and Solutions

The Hamilton operator indicates that two photons are simultaneously cre-
ated or annihilated. Find the 4 × 4 matrix M (ζ, ζ) such that

U (ζ)b1 U † (ζ)
 
b1 b1
   
† † † †
 b1  †  U (ζ)b 1 U (ζ)  = M (ζ, ζ)  b1 

U (ζ)   U (ζ) ≡  .

†
b2  U (ζ)b2 U (ζ)  b2
b†2 U (ζ)b†2 U † (ζ) b†2

Show that
cosh(s) 0 0 −eiθ sinh(s)
 
0 cosh(s) −e−iθ sinh(s) 0
M (ζ, ζ) = 
 
iθ
0 −e sinh(s) cosh(s) 0

−e−iθ sinh(s) 0 0 cosh(s)

where ζ = seiθ . Show that the determinant of the matrix is +1.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 461

Chapter 21

Trace and Partial Trace

21.1 Introduction
Let |ni be the number states (n = 0, 1, . . .) and |βi be a coherent state
(β ∈ C). The trace of a bounded linear operator  could be calculated
using the number states |ni
∞
X
tr(Â) = hn|Â|ni
n=0

or using coherent states |βi

Z
1
tr(Â) = d2 βhβ|Â|βi.
π C

The integration is over the complex plane C. One usually introduces polar
coordinates β = reıφ with r ≥ 0 and φ ∈ [0, 2π).

Let H1 , H2 be complex Hilbert spaces. Let H be their Hilbert tensor prod-

uct, i.e. H = H1 ⊗ H2 . Each Hilbert space admits at least one orthonormal
basis. Then tr2 would be the operator of taking the partial trace with re-
spect to the Hilbert space H2 . The linear operation tr2 maps states in the
Hilbert space H into states in the Hilbert space H1 . Analogously we con-
sider tr1 . For number states we could utilize |n1 i⊗|n2 i, where n1 = 0, 1, . . .
and n2 = 0, 1, . . ..

461
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 462

462 Problems and Solutions

21.2 Solved Problems

Problem 1. Show that
∞ Z
X 1
hn|Â|ni = hβ|Â|βi.
n=0
π C

Solution 1. We have
Z
1
|ni = hβ|ni|βid2 β
π C

and therefore
Z Z
1 1
hn| = hβ|nihβ|d2 β = hn|βihβ|d2 β .
π C π C

We also apply the completeness relations

Z ∞
1 X
I= |βihβ|d2 β, I= |nihn|.
π C n=0

Thus we have
∞ ∞  Z 
X X 1 2
tr(Â) = hn|Â|ni = hn| |βihβ|d β Â|ni
n=0 n=0
π C
∞
Z X
1
= hn|βihβ|Â|nid2 β
π C n=0
Z X∞ Z 
1 2
|δihδ|ni d2 δd2 β


= hn|γihγ|βid γ hβ|Â|
π 3 C n=0 C
Z Z Z ∞
1 X
= d2 βd2 γd2 δ hn|γihγ|βihδ|nihβ|Â|δi
π3 C C C n=0
∞
Z Z Z ! !
1 2 2 2
X
= d βd γd δ hδ| |nihn| |γi hγ|βi
π3 C C C n=0
Z Z Z
1
= d2 βd2 γd2 δhδ|γihδ|γihγ|βihβ|Â|δi
π3 C C C
Z Z
1
= d2 βd2 γhδ|βihβ|Â|δi
π2 C C
Z
1
= d2 δhδ|Â|δi.
π C
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 463

Trace and Partial Trace 463

Problem 2. The trace of an analytic function f (b, b† ) can be calculated

as
∞
X
tr(f (b, b† )) = hn|f (b, b† )|ni
n=0

where { |ni : n = 0, 1, 2, . . . } are the number states. A second method

consists of obtaining the normal order function of f and integrating over
the complex plane
Z
1 (n)
tr(f (b, b† )) = f (β, β ∗ )d2 β.
π C
†
(i) Find the trace of e−b b using this second method, where  > 0.
(ii) Compare with the first method.

†
Solution 2. (i) The normal order form of e−b b
is given by
∞
† X 1 −
e−b b
= (e − 1)k (b† )k bk .
k!
k=0

Thus we have to calculate the integral

Z X∞
1 1 −
(e − 1)k (β ∗ )k β k d2 β .
π C k!
k=0

We set β = reiφ . Thus ββ ∗ = r2 . Since d2 β → dφrdr with φ ∈ [0, 2π),

r ∈ [0, ∞) and
Z 2π Z ∞
2 1
dφ = 2π, re−ar dr =
0 0 2a

we obtain
† 1
tr(e−b b ) = .
1 − e−
(ii) Using the first method we find
∞ ∞
† X † X
tr(e−b b ) = hn|e−b b |ni = hn|e−n |ni
n=0 n=0
X∞ ∞
X
−n
= e hn|ni = e−n
n=0 n=0
1
= .
1 − e−
Thus the first method is simpler to apply for this case.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 464

464 Problems and Solutions

Problem 3. The single-mode squeezed state |ζi is given by |ζi = S(ζ)|0i,

where
ζ∗
 
ζ
S(ζ) = exp − (b† )2 + b2
2 2
with ζ ∈ C. One sets ζ = seiθ with s ≥ 0. Why does the squeezed state |ζi
cannot be used to calculate the trace?

Solution 3. The single-mode squeezed operator S(ζ) can be written as

   
1 † 2 iθ 1 † †
S(ζ) = exp − (b ) e tanh(s) exp − (b b + bb ) ln(cosh(s))
2 2
 
1 2 −iθ
× exp b e tanh(s) .
2
From b|0i = 0|0i and |ζi = S(ζ)|0i we find
∞ p  n
p X ((2n)!) 1
|ζi = (sech(s)) − eiθ tanh(s) |2ni.
n=0
n! 2
Consequently the single-mode squeezed state is a superposition only of even
number states. Thus the single-mode squeezed states are not complete.
Note that the coherent states are overcomplete. The coherent squeezed
states D(β)S(ζ)|0i also form an overcomplete set. One has
Z
1
d2 β|β, ζihβ, ζ| = I
π C
where I is the identity operator.

Problem 4. Consider the two-mode squeezed state

† †
|ψi = es(b1 b2 −b1 b2 ) |00i
where |00i ≡ |0i ⊗ |0i and s is the squeezing parameter. This state can also
be written as
∞
1 X
|ψi = (tanh(s))n |ni. ⊗ |ni.
cosh(s) n=0
This is the Schmidt basis for this state. The density operator ρ is given by
ρ = |ψihψ|. Calculate the partial traces using the number states.

Solution 4. We have
∞ ∞
1 X X
ρ= (tanh(s))n |ni ⊗ |ni (tanh(s))m hm| ⊗ hm|
(cosh(s))2 n=0 m=0
∞ X∞
1 X
= (tanh(s))n (tanh(s))m |nihm| ⊗ |nihm|.
(cosh(s))2 m=0 n=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 465

Trace and Partial Trace 465

Let I be the identity operator. Using that hk|ni = δkn and hm|ki = δmk
we have
∞
X
ρ1 = (I ⊗ hk|)ρ(I ⊗ |ki)
k=0
∞ ∞
1 X X
= (I ⊗ hk|) (tanh(s))n (tanh(s))m |nihm| ⊗ |nihm|(I ⊗ |ki)
(cosh(s))2 m,n=0
k=0
∞ ∞
1 X X
= (tanh(s))n (tanh(s))m |nihm|δkn δmk
(cosh(s))2
k=0 m,n=0
∞
1 X
= (tanh(s))2k |kihk|.
(cosh(s))2
k=0

We obtain the same result for ρ2 .

Problem 5. Use the reduced density operators ρ1 and ρ2 from the pre-
vious problem and calculate the entanglement

E(s) := −tr(ρ1 log2 (ρ1 )) = −tr(ρ2 log2 (ρ2 )).

Discuss E as a function of the squeezing parameter s.

Solution 5. We have
∞
!
1 X
E(s) = −tr (tanh2k (s))|kihk|
cosh2 (s) k=0
∞
!!
1 X
× log2 2 (tanh2` (s))|`ih`| .
cosh (s) `=0

The two matrices inside the trace are diagonal matrices and thus the prod-
uct is again a diagonal matrix. Thus
∞
!
1 X
2k tanh2k (s)
E(s) = − tanh (s) log2 .
cosh2 (s) k=0 cosh2 (s)

Using the property of log it follows that

∞ ∞
log2 (tanh2 (s)) X 2k log2 (cosh2 (s)) X
E(s) = − k tanh (s) + tanh2k (s).
cosh2 (s) k=0
cosh 2
(s) k=0

The identity
∞
X
tanh2k (s) ≡ cosh2 (s)
k=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 466

466 Problems and Solutions

follows from a geometric series. The identity

∞
X
k tanh2k (s) ≡ sinh2 (s) cosh2 (s)
k=0

can be obtained from the first identity by parameter differentiation with

respect to s. Using these results we obtain

E(s) = − sinh2 (s)(log2 (sinh2 (s)) − log2 (cosh2 (s)) + log2 (cosh2 (s))
= − sinh2 (s) log2 (sinh2 (s)) + (sinh2 (s) + 1) log2 (cosh2 (s))
= − sinh2 (s) log2 (sinh2 (s)) + cosh2 (s) log2 (cosh2 (s)).

For s = 0 we have E(s = 0) = 0.

Problem 6. Consider the product Hilbert space C2 ⊗H, where H denotes

an arbitrary Hilbert space. For the one-Bose system one would set H =
`2 (N0 ). An arbitrary pure state in this product Hilbert space can be written
as
|ψi = |0i ⊗ |φ0 i + |1i ⊗ |φ1 i
where |φ0 i, φ1 ∈ H and |0i, |1i forms an orthonormal basis in the Hilbert
space C2 . The condition that the state |ψi to be normalized, i.e. hψ|ψi = 1
leads to the constraint

hφ0 |φ0 i + hφ1 |φ1 i = 1.

If we assume that |φ0 i and |φ1 i have identical norms, then |ψi takes the
form
1
|ψi = √ (|0i ⊗ |ϕ0 i + |1i ⊗ |ϕ1 i)
2
where |φ0 i = √12 |ϕ0 i, |φ1 i = √12 |ϕ1 i and |ϕ0 i, ϕ1 i are normalized. Defining
the reduced density operators using the partial trace as

ρ1 = trC2 (|ψihψ|), ρ2 = trH (|ψihψ|)

the entanglement of |ψi is given by

E(|ψi) = −tr(ρ1 log2 (ρ1 )) = −tr(ρ2 log2 (ρ2 )).

Find ρ1 , ρ2 , the nonzero eigenvalues of ρ1 , ρ2 and E(|ψi).

Solution 6. Straightforward calculation yields

ρ1 = |φ0 ihφ0 | + |φ1 ihφ1 |

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 467

Trace and Partial Trace 467

and

ρ2 = hφ0 |φ0 i|0ih0| + hφ1 |φ0 i|0ih1| + hφ0 |φ1 i|1ih0| + hφ1 |φ1 i|1ih1|.

Applying the constraint hφ0 |φ0 i + hφ1 |φ1 i = 1 we find that the nonzero
eigenvalues of ρ1 and ρ2 are given by
1 p
λ= (1 + (1 − 2hφ0 |φ0 i)2 + 4|hφ0 |φ1 i|2 )
2
and 1−λ. Thus E(|ψi) = −λ log2 (λ)−(1−λ) log2 (1−λ). The entanglement
is described exclusively by hφ0 |φ0 i and |hφ0 |φ1 i|2 .

Programming Problem

Problem 1. Give an implementation of number states so that they can

utilized to calculate traces.

Solution 1.
/* numberstates.cpp */

#include <iostream>
#include "symbolicc++.h"

int main(void)
{
// b is the Bose annihilation operator, bd is the creation operator
// N[j] is the number state |j>, DN[j] is the dual state <j|
Symbolic b("b"), bd("bd"), N("N"), DN("DN"), m("m"), n("n"), x("x");
b = ~b; bd = ~bd; N = ~N; DN = ~DN;
BindingEquations rules = (b*N[0]==0,DN[0]*bd==0,
(n,b*N[n]==sqrt(n)*N[n-1]),
(n,bd*N[n]==sqrt(n+1)*N[n+1]),
(n,DN[n]*b==D[n+1]*sqrt(n+1)),
(n,DN[n]*bd==DN[n-1]*sqrt(n)),
(n,DN[n]*N[n]==1),
(m,n,DN[m]*N[n]==0));
Symbolic r1 = b*b*N[n];
cout << "r1 = " << r1.subst_all(rules) << endl;
cout << "r1(n=1) = " << r1[n==1].subst_all(rules) << endl;
cout << "DN[2]*N[3] = " << (DN[2]*N[3]).subst_all(rules) << endl;
cout << "DN[4]*N[4] = " << (DN[4]*N[4]).subst_all(rules) << endl;
Symbolic r2 = b*N[n];
cout << "r2 = " << r2.subst_all(rules) << endl;
Symbolic r3 = DN[n]*r2;
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 468

468 Problems and Solutions

cout << "r3 = " << r3.subst_all(rules) << endl;

Symbolic r4 = b*N[n+1];
cout << "r4 = " << r4.subst_all(rules) << endl;
Symbolic r5 = DN[n]*r4;
cout << "r5 = " << r5.subst_all(rules) << endl;
Symbolic r6 = b*N[n];
cout << "r6 = " << r6.subst_all(rules) << endl;
Symbolic r7 = DN[n+1]*r6;
cout << "r7 = " << r7.subst_all(rules) << end;
return 0;
}

21.3 Supplementary Problems

Problem 1. Consider the density operator

exp(−~ωb† b/kB T )
ρ= .
tr(exp(−~ωb† b/kB T ))

Show that ρ expressed with number states |ni (n = 0, 1, . . .) is given by

∞
X hnin
ρ= |nihn|
n=0
(1 + hni)n+1

where
hni := tr(ρb† b) = (exp(~ω/kB T ) − 1)−1 .

Problem 2. Let |0i, |1i be the standard basis in C2 and |β1 i, |β2 i be
coherent states. Consider the normalized state

|ψi = c0 |0i ⊗ |β1 i + c1 |1i ⊗ |β2 i

where c0 and c1 are the normalization constants. Find ρ1 = trC2 (|ψihψ|).

Problem 3. Can the squeezed state |ζi be used to calculate the trace of
an operator? Calculate Z
1
|ζihζ|dζ.
π C
Discuss.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 469

Chapter 22

Entanglement

22.1 Introduction
In the original paper of Einstein, Podolsky and Rosen the spin version of
entanglement was not used. They considered measurement of position and
momentum observables for two particles in one-dimensional motion. The
entangled state Z ∞
|ψi = |pi ⊗ | − pie−i`p/~ dp
−∞

was studied, where the first component in the tensor product refers to
particle 1 and the second to particle 2. The state |ψi is thus a superposition
of simultaneous eigenkets of the momentum operators P̂1 and P̂2 of the two
particles with associated eigenvalues p and −p, respectively. Thus |ψi is
itself an eigenket of
P̂1 ⊗ I + I ⊗ P̂2
with the eigenvalue 0. The entangled state |ψi is also an eigenket of the
operator
Q̂1 ⊗ I + I ⊗ Q̂2
where Q̂1 and Q̂2 are the position operators of the two particles. The
maximally entangled state of the original EPR pair can also be written as
Z
1
|ψi = dk exp(ik · r1 ) exp(ik · r2 ) = δ(r1 + r2 )
(2π)3/2

469
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 470

470 Problems and Solutions

where δ denotes the Dirac delta function and · is the scalar product. If
the potential energy between the two particles is assumed to be spherically
symmetric, we can assume that the vectors r1 and r2 lie in the xy-plane in
the laboratory frame without loss of generality.

Entangled state are the Schrödinger cat states

1
p (|βi ⊗ | − βi + | − βi ⊗ |βi)
Nβ

where |βi denotes a coherent state. For squeezed states |ζi we can consider
the entangled state

1
p (|ζi ⊗ | − ζi + | − ζi ⊗ |ζi).
Nζ

The two-mode state

† †
|ψi = es(b1 b2 −b1 b2 ) |0i ⊗ |0i
∞
1 X
= (tanh(s))n |ni ⊗ |ni
cosh(s) n=0

where s is the squeezing parameter and |ni are the number states is an
entangled state. Let |0i, |1i be an orthonormal basis in C2 and |βi be a
coherent state. Then

1
√ (|0i ⊗ |βi + |1i ⊗ | − βi)
N

is an entangled state. One can also consider

1
√ (|0i ⊗ |ζi + |1i ⊗ | − ζi)
N

where |ζi is a squeezed state.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 471

Entanglement 471

22.2 Solved Problems

Problem 1. Let
 2
N N
2 X 1 X
f (p, q) := qj  + (pj − pk )2
N j=1 N
j,k=1

and  2
N N
2 X 1 X
g(p, q) := pj  + (qj − qk )2 .
N j=1 N
j,k=1

The Wigner function of the pure entangled N -mode state is given by

 N
2
exp −e−2s f (p, q) − e2s g(p, q)


W (q, p) = (1)
π
where q = (q1 , q2 , . . . , qN ) and p = (p1 , p2 , . . . , pN ) are the positions and
momenta of the N modes and s is the squeezing parameter with equal
squeezing in all initial modes. Consider the case N = 2. What happens if
s → ∞?

Solution 1. The state W (q, p) is always positive, symmetric among

the N modes and becomes peaked at qi − qj = 0 (i, j = 1, 2, . . . , N ) and
p1 + p2 + · · · + pn = 0 for large squeezing parameter s. From (1) we have

W (q1 , q2 , p1 , p2 ) =
4  
2


2
exp −e−2s (q1 + q2 ) + (p1 − p2 )2 − e2s (p1 + p2 )2 + (q1 − q2 )2 .
π
For s → ∞ we find in the sense of generalized functions

Cδ(q1 − q2 )δ(p1 + p2 )

where δ denotes the Dirac delta function. This makes a connection to the
original EPR state of Einstein, Podolsky and Rosen. Thus for large s the
function W peaks at q1 − q2 = 0 and p1 + p2 = 0.

Problem 2. Consider the operator

† †
U (r) := e−r(b1 b2 −b1 b2 )

where b†1 , b†2 are Bose creation operators and b1 , b2 are Bose annihilation
operators and r ∈ R. Thus b†1 = b† ⊗ I, b†2 = I ⊗ b† . Let |0i ⊗ |0i be the
vacuum state, i.e.

(b ⊗ I)(|0i ⊗ |0i) = 0|0i ⊗ |0i, (I ⊗ b)(|0i ⊗ |0i) = 0|0i ⊗ |0i.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 472

472 Problems and Solutions

(i) Calculate |ψ(r)i = U (r)(|0i ⊗ |0i).

(ii) Let

X̂1 := b1 + b†1 = b ⊗ I + b† ⊗ I, Ŷ1 := −i(b1 − b†1 ) = −i(b ⊗ I − b† ⊗ I),

X̂2 := b2 + b†2 = I ⊗ b + I ⊗ b† , Ŷ2 := −i(b2 − b†2 ) = −i(I ⊗ b − I ⊗ b† ).

Find var(X̂1 + X̂2 ), var(Ŷ1 − Ŷ2 ), where

var(A) := hA2 i − hAi2

is the variance.
(iii) What happens in the limit r → ∞ to the state |ψ(r)i?

Solution 2. (i) We find

p ∞
X
|ψ(r)i = U (r)(|0i ⊗ |0i) = (1 − λ2 ) λn |ni ⊗ |ni
n=0
√
where λ = tanh(r) and therefore 1 − λ2 = 1/ cosh(r). The entanglement
of this state can be viewed as an entanglement between quadrature phases in
the two modes (EPR entanglement) or as an entanglement between number
and phase in the two modes.
(ii) We find

var(X̂1 + X̂2 ) = 2e−2r , var(Ŷ1 − Ŷ2 ) = 2e−2r .

(iii) The state |ψ(r)i approaches a simultaneous eigenstate of X̂1 + X̂2 and
Ŷ1 − Ŷ2 .

Problem 3. Consider a quantum-mechanical system governed by the

Hamilton operator

Ĥ = ~ω1 b†1 b1 + ~ω2 b†2 b2 + ~χb†1 b1 b†2 b2

where b1 and b2 are Bose annihilation operators for two distinct harmonic
oscillator modes, respectively and χ is a coupling constant. Such a Hamil-
ton operator for optical systems describes a four-wave mixing process, when
the constant χ is then proportional to the third order susceptibility. It can
also be used to describe two distinct modes interaction in Bose conden-
sate. Furthermore, it describes the effective interaction of output pump
and probe fields of an optical-cavity mediated by a two-level atom, in the
dispersive limit. Let
|ψ(t = 0)i := |β1 i ⊗ |β2 i
where |β1 i and |β2 i are coherent states.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 473

Entanglement 473

(i) Find U (t)|ψ(t = 0)i, where

U (t) = exp(−iĤt/~).

(ii) Consider the special case t = π/χ. Discuss.

(iii) Consider the four cases a) ω1 = 2χ, ω2 = 2χ, b) ω1 = 2χ, ω2 = χ,
c) ω1 = χ, ω2 = 2χ, d) ω1 = χ, ω2 = χ for |ψ(π/χ)i.

Solution 3. (i) We find

∞
2 X (β1 e−iω1 t )m
|ψ(t)i = U (t)|β1 i⊗|β2 i = e−|β1 | /2
√ |mi⊗|β2 e−iω2 t e−iχmt i.
m=0
m

(ii) For t = π/χ we have


1 m = even
exp(−iχmt) = exp(−imπ) =
−1 m = odd

Thus

|ψ(π/χ)i = |β1+ e−iπω1 /χ i ⊗ |β2 e−iπω2 /χ i + |β1− e−iπω1 /χ i ⊗ |−β2 e−iπω2 /χ i

or

|ψ(π/χ)i = |β1 e−iπω1 /χ i ⊗ |β2+ e−iπω2 /χ i + |−β1 e−iπω1 /χ i ⊗ |β2− e−iπω2 /χ i

where
1
|± e−iπωk χ i :=(|e−iπωk /χ i ± |−e−iπωk /χ i)
2
with k = 1, 2 and  = β1 , β2 . Hence the state is entangled.
(iii) For case a) we find

|Φ+ i = |β1 i ⊗ |β2+ i + |−β1 i ⊗ |β2− i.

For case b) we find

|Φ− i = |β1 i ⊗ |β2+ i − |−β1 i ⊗ |β2− i.

For case c) we find

|Ψ+ i = |β1 i ⊗ |β2− i + | − β1 i ⊗ |β2+ i.

For case d) we find

|Ψ− i = |β1 i ⊗ |β2− i − | − β1 i ⊗ |β2+ i.

These states may be considered as Bell states. However these states are not
perfectly orthogonal, but for large-amplitude fields |β1 |, |β2 |  1 this can
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 474

474 Problems and Solutions

be achieved approximately. Furthermore there is an asymmetry in these

states.

Problem 4. Discuss the entanglement of the state

δ(ωs + ωi − ωp )δ(ks + ki − kp )b†s (ω(ki ))b†i (ω(ki ))|0i

XX
|Ψi =
s i

which appears with spontaneous parametric down conversion. Here ωj , kj

(j = s, i, p) are the frequencies and wave vectors of the signal (s), idler (i),
and pump (p) respectively, ωp and kp can be considered as constants while
b†s and b†i are the respective Bose creation operators for the signal and idler.

Solution 4. The entanglement of this state can be thought of as the

superposition of an infinite number of two-photon states, corresponding to
the infinite number of ways the spontaneous parametric down conversion
signal-idler can satisfy the expression for energy and momentum conserva-
tion (owing to the delta functions)

~ωs + ~ωi = ~ωp , ~ks + ~ki = ~kp .

Even if there is no precise knowledge of the momentum for either the signal
or the idler, the state does give precise knowledge of the momentum corre-
lation of the pair. In EPR’s language, the momentum for neither the signal
photon nor the idler photon is determined. However, if measurement on
one of the photons yields a certain value, then the momentum of the other
photon is determined.

Problem 5. Consider the function

 
1 1 2 2s 1 2 −2s
G(x1 , x2 ; r) = √ exp − (x1 + x2 ) e − (x1 − x2 ) e
2π 4 4

where s > 0 is the squeezing parameter. Find

lim (G(x1 , x2 ; s), φ(x1 , x2 ))

s→∞

in the sense of generalized functions, where φ ∈ S(R2 ). Here S(R2 ) is

the set of all infinitely-differentiable functions which decrease as |x| → ∞,
together with all their derivatives, faster than any power of |x|−1 .

Solution 5. We find

lim (G(x1 , x2 ; s), φ(x1 , x2 )) → φ(x1 , x1 ).

r→∞
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 475

Entanglement 475

Thus
lim G(x1 , x2 ; s) → δ(x1 − x2 )
s→∞

in the sense of generalized functions, where δ is the Dirac delta function.

Problem 6. Consider the operator

U = exp((b†1 b2 − b†2 b1 )),  ∈ R.

(i) Find U† b1 U , U† b2 U . Consider the special case  = π/4.

†
(ii) Find D = Uπ/4 (b†1 b1 − b†2 b2 )Uπ/4 .
(iii) Solve the eigenvalue problem D|δi = d|δi.

Solution 6. (i) Using the expansion

1 1
e B̂e− = B̂ + [Â, B̂] + [Â, [Â, B̂]] + · · · + [Â, [Â, · · · , [Â, B̂] . . .]] + · · ·
2 n!
we find

U† b1 U = b1 cos() + b2 sin(), U† b2 U = −b1 sin() + b2 cos().

For the special case  = π/4 we obtain

† 1 † 1
Uπ/4 b1 Uπ/4 = √ (b2 + b1 ), Uπ/4 b2 Uπ/4 = √ (b2 − b1 )
2 2
√
since sin(π/4) = cos(π/4) = 1/ 2.
(ii) From (ii) we find

U† b†1 U = b†1 cos() + b†2 sin(), U† b†2 U = −b†1 sin() + b†1 cos().

Thus
D = b†1 b2 + b1 b†2 .
(iii) The eigenvalue problem D|δi = d|δi can be rewritten as

(b†1 b1 − b†2 b2 )|νi = d|νi

where
|νi = Uπ/4 |δi.
The eigenvalue problem can easily be solved since b† b|ni = n|ni. We find

 |n + di ⊗ |ni d ∈ Z+
(n)
|ν i = |ni ⊗ |ni d=0
|ni ⊗ |n − di d ∈ Z−

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 476

476 Problems and Solutions

where Z+ denotes the positive set of integers and Z− denotes the negative
set of integers. The eigenvalue d has countable degeneracy corresponding
to the one-integer parameter set |ν (n) i of eigenstates. In order to solve for
the original eigenvalues we have to compute their transformation under the
action of the operator Uπ/4 , i.e.

†
|δ (n) i = Uπ/4 |ν (n) i.

We consider the Schwinger two-bosons realization of the su(2) Lie algebra

1 †
J+ := b1 b†2 , J− := b†1 b2 , J3 := (b b2 − b†1 b1 )
2 2
with [J+ , J− ] = 2J3 , [J3 , J± ] = ±J± . Thus
π 
Uπ/4 = exp (J+ − J− ) .
4
Using the Baker-Campbell-Hausdorff formula we find

¯ − ) = exp(ηJ+ ) exp(βJ3 ) exp(−η̄J− )

exp(ξJ+ − ξJ

where
ξ
η= tan(ξ), β = ln(1 + |η|2 ).
|ξ|
Thus
Uπ/4 = exp(b1 b†2 ) exp(ln 2(b†2 b2 − b†1 b1 )) exp(−b†1 b2 )
where we used that tan(π/4) = 1.

Problem 7. Consider the Hamilton operator

N −1 N −1
hij b†i bj + Vijlm b†i b†j bm bl .
X X
Ĥ =
i,j=0 i,j,l,m=0

The operators b†j are Bose creation operators and the operators bj are Bose
annihilation operators. Let
n̂j := b†j bj
be the particle number operator of mode j for an appropriate basis. Show
that an eigenstate |ψi of Ĥ is entangled or

[Ĥ, n̂j ]|ψi = 0|ψi

i.e. we have eigenvalue equation with eigenvalue 0.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 477

Entanglement 477

Solution 7. Suppose |ψi is not entangled. We write |ψi as

|ψi = |n0 i ⊗ · · · ⊗ |nN −1 i

where { |0j i, |1j i, . . . } is a basis for particles in mode j with 0 ≤ j < N .

We define the creation, annihilation and number operators, for particles in
mode j, by
√
b†j |nj i := nj + 1|n + 1j i,
p
bj |nj i := nj |n − 1j i

and n̂j := b†j bj . We have the eigenvalue equation for |ψi, namely Ĥ|ψi =
λ|ψi. Thus

[Ĥ, n̂i ]|ψi = Ĥ n̂i |ψi − n̂i Ĥ|ψi = Ĥni |ψi − λn̂i |ψi
= ni Ĥ|ψi − λn̂i |ψi = λni |ψi − λni |ψi
= 0|ψi.

Problem 8. Let |βi be a coherent state. Consider the entangled coherent

state
|ψi = C(|β1 i ⊗ |β2 i + eiφ |−β1 i ⊗ |−β2 i)
where C is the normalization factor and φ ∈ R.
(i) Find the normalization factor C.
(ii) Calculate the partial trace using the basis { |ni ⊗ I : n = 0, 1, 2, . . . }
where { |ni : n = 0, 1, 2, . . . } are the number states and I is the identity
operator.

Solution 8. (i) Since

 
1
hβ|γi = exp − (|β|2 + |γ|2 ) + βγ ∗
2

for coherent states |βi and |γi, we have

hβ|βi = 1, hβ|−βi = exp(−2|β|2 ) .

We find from the condition hψ|ψi = 1 that

1 = |C|2 (2 + 2 cos(φ) exp(−2|β1 |2 − 2|β2 |2 )).

Thus
1
C=p .
2 + 2 cos(φ) exp(−2|β1 |2 − 2|β2 |2 )
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 478

478 Problems and Solutions

(ii) We have to calculate

∞
X
tr1 (|ψihψ|) = C 2 ((hn| ⊗ I)(|β1 i ⊗ |β2 i + eiφ |−β1 i ⊗ |−β2 i)
n=0
×(hβ1 | ⊗ hβ2 | + e−iφ hβ1 −| ⊗ hβ2 −|)(|ni ⊗ I)).
Thus
∞
X
tr1 (|ψihψ|) = C 2 (hn|β1 ihβ1 |ni|β2 ihβ2 | + e−iφ hn|β1 ihβ1 −|ni|β2 ihβ2 −|
n=0
+eiφ hn|−β1 ihβ1 |ni|−β2 ihβ2 | + hn|−β1 ihβ1 −|ni|−β2 ihβ2 −|).
Using
2
e−|β1 | (|β1 |2 )n
hn|β1 ihβ1 |ni =
n!
−|β1 |2
e (−|β1 |2 )n
hn|β1 ihβ1 −|ni =
n!
−|β1 |2
e (−|β1 |2 )n
hn|−β1 ihβ1 |ni =
n!
−|β1 |2
e (|β1 |2 )n
hn|−β1 ihβ1 −|ni =
n!
and
∞ ∞
X (|β|2 )n 2 X (−|β|2 )n 2
= e|β| , = e−|β|
n=0
n! n=0
n!
we arrive at
2
tr1 (|ψihψ|) = C 2 (|β2 ihβ2 | + eiφ e−2|β1 | |β2 ihβ2 −|
2
+e−iφ e−2|β1 | |−β2 ihβ2 | + |−β2 ihβ2 −|).

Problem 9. A beam splitter is a simple device which can act to entangle

output optical fields. The input field described by the Bose annihilation
operator b1 is superposed on the other input field with Bose annihilation
operator b2 by a lossless symmetric beam splitter with amplitude reflec-
tion and transmission coefficients r and t. The output field annihilation
operators are given by
eb1 = B̂b1 B̂ † , eb2 = B̂b2 B̂ †

where the beam splitter operator is

 
θ † iφ
B̂ := exp (b1 b2 e − b1 b†2 e−iφ )
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 479

Entanglement 479

with the amplitude reflection and transmission coefficients

t := cos(θ/2), r := sin(θ/2).

The beam splitter gives the phase difference φ between the reflected and
transmitted fields.
(i) Assume that the input states are two independent number states |n1 i ⊗
|n2 i, where n1 , n2 = 0, 1, 2, . . .. Calculate the state B̂(|n1 i ⊗ |n2 i).
(ii) Consider the special case n1 = 0 and n2 = N .

Solution 9. (i) We obtain the state

∞
X ∞
X
B̂(|n1 i ⊗ |n2 i) = (hm1 | ⊗ hm2 |)B̂(|n1 i ⊗ |n2 i)|m1 i ⊗ |m2 i
m1 =0 m2 =0
X∞ X∞
= Bnm11nm2 2 |m1 i ⊗ |m2 i
m1 =0 m2 =0

where
n1 X
X n2
Bnm11nm2 2 = e−iφ(n1 −m1 ) (−1)n1 −k rn1 +n2 −k−` tk+`
k=0 `=0
√
n1 !n2 !m1 !m2 !
× δm ,n +k−` δm2 ,n1 −k+`
k!(n1 − k)!`!(n2 − `)! 1 2

with δm,n is the Kronecker delta. When the total number of input photons
is N = n1 + n2 , the output state becomes an (N + 1)-dimensional entangled
state.
(ii) We obtain from the results of (i)

N
X
B(|0i ⊗ |N i) = cN
k |ki ⊗ |N − ki
k=0

where the expansion coefficients are given by

 1/2
N
cN
k = rk tN −k eikφ .
k

Problem 10. The beam splitter operator is given by

 
θ † iφ
B̂ = exp (b1 b2 e − b1 b†2 e−iφ ) .
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 480

480 Problems and Solutions

The input field described by the Bose operator b1 is superposed with another
input field with Bose operator b2 by a lossless symmetric beam splitter with
amplitude reflection and transmission coefficients r and t, i.e.

t := cos(θ/2), r := sin(θ/2).

The output-field Bose annihilation operators are given by

eb1 = B̂b1 B̂ † , eb2 = B̂b2 B̂ † .

Consider the input state (product state of two independent Fock states)

|n1 , n2 i ≡ |n1 i|n2 i ≡ |n1 i ⊗ |n2 i

where n1 , n2 = 0, 1, . . .. Calculate B̂|n1 , n2 i. Is this state entangled?

Solution 10. We obtain

∞
X ∞
X ∞
X ∞
X
B̂|n1 , n2 i = |N1 , N2 ihN1 , N2 |B̂|n1 , n2 i = BnN11nN22 |N1 , N2 i
N1 =0 N2 =0 N1 =0 N2 =0

where
n1 X
X n2
BnN11nN22 = e−iφ(n1 −N1 ) (−1)n1 −k rn1 +n2 −k−l tk+l
k=0 l=0
√
n1 !n2 !N1 !N2 !
× δN ,n +k−l δN2 ,n1 −k+l
k!(n1 − k)!l!(n2 − l)! 1 2

where δ is the Kronecker delta function. When the total number of input
photons is N = n1 +n2 with N ≥ 1, the output state is an (N +1) entangled
state.

Problem 11. Consider the product Hilbert space C2 ⊗H where H denotes

an arbitrary Hilbert space. For the one-Bose system we would set H =
l2 (N). An arbitrary pure state in this product Hilbert space can be written
as
|ψi := |0i ⊗ |φ0 i + |1i ⊗ |φ1 i
where |φ0 i, |φ1 i ∈ H and { |0i, |1i } forms an orthonormal basis in C2 . The
condition for the state |ψi to be normalized, i.e. hψ|ψi = 1, leads to the
constraint hφ0 |φ0 i + hφ1 |φ1 i = 1. If we assume that |φ0 i and |φ1 i have
identical norms, then |ψi takes the form

1
|ψi = √ (|0i ⊗ |ϕ0 i + |1i ⊗ |ϕ1 i)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 481

Entanglement 481

where |φ0 i = √12 |ϕ0 i, |φ1 i = √12 |ϕ1 i and |ϕ0 i, |ϕ1 i are normalized. Defin-
ing the reduced density matrices (using the partial trace)

ρ1 := trC2 (|ψihψ|), ρ2 := trH (|ψihψ|)

the entanglement of |ψi is given by

E(|ψi) := −tr(ρ1 log2 (ρ1 )) = −tr(ρ2 log2 (ρ2 )).

Describe the entanglement.

Solution 11. Straightforward calculation yields

ρ1 = |φ0 ihφ0 | + |φ1 ihφ1 |

and

ρ2 = hφ0 |φ0 i|0ih0| + hφ1 |φ0 i|0ih1| + hφ0 |φ1 i|1ih0| + hφ1 |φ1 i|1ih1|.

Applying the constraint we find that the non-zero eigenvalues of ρ1 and ρ2

are given by
1 p 
λ(hφ0 |φ0 i, |hφ0 |φ1 i|2 ) := 1 + (1 − 2hφ0 |φ0 i)2 + 4|hφ0 |φ1 i|2
2
and 1 − λ. Thus

E(|ψi) = −λ log2 (λ) − (1 − λ) log2 (1 − λ).

The entanglement is described exclusively by hφ0 |φ0 i and |hφ0 |φ1 i|2 . Fur-
thermore we have the inequality
1
|hφ0 |φ1 i|2 ≤ hφ0 |φ0 i − hφ0 |φ0 i2 ≤ .
4

Problem 12. Consider the superposition (macroscopic quantum super-

position states) |φ0 i = c0 (|αi + |−αi) and |φ1 i = c1 (|βi + |−βi) where
c0 , c1 ∈ C and |αi, |−αi, |βi, |−βi, are coherent states, i.e.

|ψi = c0 |0i ⊗ (|αi + |−αi) + c1 |1i ⊗ (|βi + |−βi).

Discuss the entanglement.

Solution 12. The conditions for entanglement from the previous problem
can be applied. In this case the normalization condition yield
2 2
2|c0 |2 (1 + e−2|α| ) + 2|c1 |2 (1 + e−2|β| ) = 1.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 482

482 Problems and Solutions

2
Consequently hφ0 |φ0 i = 2|c0 |2 (1 + e−2|α| ) and
1 2 1 2
2 (e− 2 |α−β| + e− 2 |α+β| )2
|hφ0 |φ1 i| = hφ0 |φ0 i(1 − hφ0 |φ0 i) .
(1 + e−2|α|2 )(1 + e−2|β|2 )
It is convenient to define the real valued quantities
1 2 1 2
2 −2|α|2 (e− 2 |α−β| + e− 2 |α+β| )2
p00 := hφ0 |φ0 i = 2|c0 | (1 + e ), p01 := .
(1 + e−2|α|2 )(1 + e−2|β|2 )
Thus we obtain
|hφ0 |φ1 i|2 = p00 (1 − p00 )p01 .
The maximum entanglement occurs when hφ0 |φ1 i = 0 and hφ0 |φ0 i = 12 .
6 0 for hφ0 |φ0 i = 21 , the maximum
Since the above equation implies hφ0 |φ1 i =
entanglement is approached asymptotically for α = 0, |β| → ∞ or β = 0,
|α| → ∞. This is due to

|α − β|2 = |α|2 + |β|2 − 2<(αβ), |α + β|2 = |α|2 + |β|2 + 2<(αβ) .

In other words, for α, β 6= 0 one term shrinking in the numerator of p01

implies the other is growing. To find the entanglement we first determine
the eigenvalues of ρ1 and ρ2 which are now given by
1 p 
λ= 1 + 1 − 4(1 − p01 )p00 (1 − p00 ) .
2

Problem 13. Consider the case when |φ0 i is described by a number state
and |φ1 i is described by a coherent state, i.e.

|ψi = c0 |0i ⊗ |ni + c1 |1i ⊗ |αi.

The scalar product between a number state |ni and a coherent state |αi is
given by
2 α
n
1
hn|αi = e− 2 |α| √ .
n!
Discuss the entanglement.

Solution 13. The normalization condition for this case gives

|c0 |2 + |c1 |2 = 1 .

Consequently hφ0 |φ0 i = |c0 |2 , and

2 |α|2n
|hφ0 |φ1 i|2 = |c0 |2 (1 − |c0 |2 )e−|α| .
n!
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 483

Entanglement 483

It is again convenient to define the quantities p00 and p01 as

2 |α|2n
p00 := hφ0 |φ0 i = |c0 |2 , p01 := e−|α| .
n!

Thus we obtain |hφ0 |φ1 i|2 = p00 (1 − p00 )p01 . The entanglement can again
be determined from p00 and p01 and proceeds as described in the previous
problem.

Problem 14. Let |βi be a coherent state. Consider the entangled state

1
|ψi = p (|βi ⊗ |βi − | − βi ⊗ | − βi).
Nβ

(i) Find the normalization Nβ .

(ii) Express the entangled state |ψi using Fock states |ni ⊗ |mi with n, m ∈
N0 .

Solution 14. Using hβ| − βi = exp(−2|β|2 ) we obtain

Nβ = 2 − 2 exp(−4|β|2 ).

(ii) We obtain
∞
2 exp(−|β|2 ) X β n+m
|ψi = p √ |ni ⊗ |mi .
Nβ n,m | n+m odd
n!m!

Thus the total number of photons is always odd.

Problem 15. Let |βi, |γi be coherent states. Consider the balanced
entangled coherent state

1
|ψi = √ (|βi ⊗ |γi + eiφ | − βi ⊗ | − γi)
N

and φ = π/2. Consider the operator

U = exp(iπ(b† b ⊗ I + I ⊗ b† b)).

Find U |ψi and U 2 |ψi. Discuss.

Solution 15. Since eiπ = −1 and eiπ/2 = i and

exp(iφb† b)|βi = |βeiφ i

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 484

484 Problems and Solutions

we find the entangled state

i
U |ψi = √ (|βi ⊗ |γi − i| − βi ⊗ | − γi).
2

Applying the operator U again yields U 2 |ψi = |ψi. Note that

† †
eiπb b be−iπb b
= −b.

Problem 16. Consider the unitary evolution operator for the beam split-
ter
UBS = exp(θ(b†1 b2 eiφ − b1 b†2 e−iφ ))
where the real angular parameter θ determines the transmission and reflec-
tion coefficients via T = t2 = cos2 (θ) and R = r2 = sin2 (θ). The internal
phase shift φ between the reflected and transmitted modes is given by the
beam splitter itself. To control φ we can place a phase shifter in one of the
output channels.
(i) Let |ψin i = |0i ⊗ |1i, i.e. the one input is a one-photon state and the
other the vacuum state. Calculate |ψout i = UBS |ψin i.
(ii) To test quantum non locality of the state |ψout i we apply the displaced
parity operator based on joint parity measurements

Π̂12 (β1 , β2 ) := D1 (β1 )D2 (β2 ) exp(iπ(n̂1 + n̂2 ))D1† (β1 )D2† (β2 )

where D1 (β1 ) and D2 (β2 ) are the unitary displacement operators. Calcu-
late
Π12 (β1 , β2 ) := hψout |Π̂12 (β1 , β2 )|ψout i.
(iii) The two mode Bell function B(β1 , β2 ) can be written as

B(β1 , β2 ) = Π12 (0, 0) + Π12 (β1 , 0) + Π12 (0, β2 ) − Π12 (β1 , β2 ).

For local realistic theory B(β1 , β2 ) should satisfy the Bell-CHSH inequality

|B(β1 , β2 )| ≤ 2.

The violation of this inequality indicates quantum non locality of the single
photon entangled state. Calculate B(β1 , β2 ) and discuss the case where
|β1 |2 = |β2 |2 .

Solution 16. (i) We find

|ψout i = t|0i ⊗ |1i + reiφ |1i ⊗ |0i.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 485

Entanglement 485

(ii) We obtain
Π(β1 , β2 ) = hψout |Π̂(β1 , β2 )|ψout i
= (4|re−iφ β1 + tβ2 |2 − 1) exp(−2(|β1 |2 + |β2 |2 )).
(iii) Let |β1 |2 = |β2 |2 = J and let γ12 be an arbitrary phase space difference
between the two coherent displacements β1 and β2 . Then we can write
β2 = β1 eiγ12 . Thus the two-mode Bell function is given by
B(β1 , β2 ) = −1 + (4J − 2)e−2J − (4J − 1)e−4J − 8rtJe−4J cos(∆)
where ∆ := γ12 + φ. When γ12 = −φ we obtain the maximal value |B|max
of the two mode Bell function
|B|max = 1 + (4J − 1)e−4J + 8rtJe−4J − (4J − 2)e−2J .

Problem 17. There are various ways in which photons can be entangled.
The photon is a spin-1 particle. One has a) polarization entanglement, b)
momentum (direction) entanglement c) time-energy entanglement d) or-
bital angular momentum states entanglement. Describe the different types
of entanglement for photons. Parametric down conversion can produce
photons that are entangled both in polarization and in space.

Solution 17. Polarization entanglement. The highest contrast in exper-

iments can be achieved for polarization-entangled states created by para-
metric down-conversion. Type-II sources can produce polarization entan-
glement directly. Parametric down-conversion or spontaneous parametric
fluorescence is the spontaneous reverse process of second-harmonic gener-
ation, or more generally speaking three-wave mixing in nonlinear optical
media (for example Beta-Barium Borate crystal). In nonlinear optics the
polarization P depends nonlinear on the electric field E. The nonlinearity
is given by a power series expansion of the polarization vector (summation
convention is used)
(1) (2) (3)
Pi = χij Ej + χijk Ej Ek + χijkl Ej Ek El + · · ·
(1)
where χij describes the normal refractive properties of a material includ-
(2)
ing any kind of birefringence. χijk is the coefficient tensor for three-wave,
because two E terms can lead to another P term, mixing in strongly non-
(3)
linear material. The χijkl term describes effects that occur at even higher
intensities, e.g. Kerr-lensing or phase conjugation.
In down-conversion one has a high-frequency pump field and two lower fre-
quency down-converted fields. Let ω be the frequency. Energy conservation
yields
~ωp = ~ω1 + ~ω2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 486

486 Problems and Solutions

whereas phase matching is given by

~kp = ~k1 + ~k2

where k is the wave vector. As most nonlinear media are birefringent

the second criterion can be satisfied by choosing an appropriate cut, thus
managing the refractive indices and therefore wave velocities such that light
can be emitted in specific directions. In type-II spontaneous parametric
down-conversion a pump beam is incident on a nonlinear optical crystal
in which pump photons spontaneously split with a low probability into
two orthogonally polarized photons called signal and idler. From energy
and momentum conversation one finds that the wavelength and emission
directions of the down conversion photons are correlated. They depend
on the pump wavelength as well as on the angle between optical axis of
the crystal and the pump beam. In the degenerate case (signal and idler
having the same wavelength) the photons leave the crystal symmetrically
with respect to the pump beam along two cones. For certain orientations
of the optical crystal, the two emission cones intersect and the photons
emerging along the intersection directions can not be assigned to one of
the two orthogonally polarized cones anymore and thus form a polarization
entangled pair. The polarization entangled state (Bell state) is described
by
1
√ (|hi ⊗ |vi + eiφ |vi ⊗ |hi)
2
where h and v denote horizontal and vertical polarizations of light. The
state is fully entangled (Bell state). By using only standard optical elements
in one of the two output beams, one can transform any one of the Bell states
into any of the other.
The energy-entangled states from down-conversion photons are the most
universal. They are present for any pair of photons. Since there are many
ways to partition the energy of the pump photon, each daughter photon
has a broad spectrum, and hence a narrow wave packet in time. The sum
of the two daughter photons energies is well-defined. They must add up to
the energy of the monochromatic pump laser photon. This correlation is
given by the energy entangled state
Z Ep
|ψi = A(E)|Eis ⊗ |Ep − Eii dE
0

where each ket describes the energy of one of the photons, s and i denotes
the signal and idler, respectively, and A(E) is the spectral distribution of
the collected down conversion light.
Another entanglement from the parametric down-conversion process is the
momentum direction entanglement. From the emission of a parametric
down-conversion source two pairs of spatial (momentum direction) modes
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 487

Entanglement 487

are extracted by pinholes. Photon pairs are emitted such that whenever
a photon is emitted into one of the inner two modes its partner will be
found in the opposite outer mode due to the phase matching in the crys-
tal. The superposition of the two inner and the two outer modes on the
beam-splitter serves to measure coincidence rates in various superposition
of the initial spatial modes. After the beam-splitters we cannot distinguish
the upper two modes from the lower two and therefore interference will be
observed in the various coincidence rates.

In a discrete version (time-bin entanglement) of energy-time entangled

states sources one sends a double pulse through the down-conversion crys-
tal. If the delay between the two pump pulses equals the time difference
between the short and the long arms of the Mach-Zehnder interferometer
then again there are two indistinguishable ways of obtaining a coincidence
detection.

22.3 Supplementary Problems

Problem 1. Consider the number states |ni (n = 0, 1, 2, . . .). Are the

states
1
√ (|ni ⊗ |ni + |n + 1i ⊗ |n + 1i)
2
1
√ (|ni ⊗ |n + 1i + |n + 1i ⊗ |ni)
2
entangled?

Problem 2. Let |ζi be a squeezed state. Is the state

1
√ (|ζi ⊗ | − ζi + | − ζi ⊗ |ζi)
N

entangled?

Problem 3. Let |ni be a number state and |βi be a coherent state. Is

the state
1
√ (|ni ⊗ |βi + |ni ⊗ | − βi)
N
entangled? Does it depend on n ∈ N0 and β ∈ C?
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 488

488 Problems and Solutions

Problem 4. Consider the coherent state |βi and the squeezed state |ζi
state. Is the state
1
√ (|βi ⊗ |ζi + |ζi ⊗ |βi)
N
entangled? Here N is a normalization factor.

Problem 5. Let

b†0 = b† ⊗ I ⊗ I, b†1 = I ⊗ b† ⊗ I, b†2 = I ⊗ I ⊗ b†

be Bose creation operators. Let |ζi (ζ ∈ C) be a squeezed state with input

mode b†0 and |β1 i, |β2 i (β1 , β2 ∈ C) be coherent states with input modes
b†1 , b†2 , respectively. Show that the state

|ζ, β1 , β2 i =
∞
−(|β1 |2 +|β2 |2 )/2
X (b† )n (β1 b†1 )n1 (β2 b†2 )n2
e Cn (ζ) √0 |0i0 ⊗ |0i1 ⊗ |0i2
n=0,n1 =0,n2 =0 n! n1 ! n2 !

is entangled. Here Cn (ζ) is the coefficient of the squeezed state with squeez-
ing parameter seiθ and is equal to 0 for all odd values of n and for n even
given by
√  n/2
n! 1
Cn (ζ) = p − eiθ tanh(s) , n even.
cosh(s)((n/2)!) 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 489

Chapter 23

Continuous Variable
Teleportation

23.1 Introduction
Quantum continuous variables provide a new approach to quantum infor-
mation processing and quantum communication. They describe highly ex-
cited quantum systems such as multi-photon fields of light. Continuous vari-
ables offer additional advantages over the single-photon system. They in-
volve the use of highly efficient telecommunication photodiodes. The coher-
ent sources of continuous entanglement are also orders of magnitude more
efficient than the spontaneous sources of discrete entanglement. Teleporta-
tion schemes can be demonstrated involving bright light sources. Entangled
states build from coherent states |βi are utilized and also Schrödinger cat
states
1
√ (|βi + | − βi).
N

23.2 Solved Problems

Problem 1. Consider the numbers states |ni with n = 0, 1, 2, . . .. Let

|Ψin i = α|0i1 + β|1i1 + γ|2i1 (1)

489
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 490

490 Problems and Solutions

be an input quantum state in mode 1. An ancilla quantum state

|Ψancilla i = |0i2 |0i3 ≡ |0i2 ⊗ |0i3

in the vacuum state is also given. The modes 2 and 3 pass through a first
parametric amplifier whose transformation is given by the operator (θ1 ∈ R)

U23 (θ1 ) = exp(θ1 (b†2 b†3 − b2 b3 )).

(i) Calculate

|Ψ0ancilla i = U23 (θ1 )|Ψancilla i = U23 (θ1 )(|0i2 ⊗ |0i3 )

and set γ1 = tanh2 (θ1 ).

(ii) Consider the product state

|Ψin i|Ψ0ancilla i ≡ |Ψin i ⊗ |Ψ0ancilla i.

The output mode 2 of the first parametric amplifier and the mode 1 are
used as the input modes of the second parametric amplifier (θ2 ∈ R)

U12 (θ2 ) = exp(θ2 (b†1 b†2 − b1 b2 )) ⊗ I3 .

Calculate
|Ψout i = U12 (θ2 )(|Ψin i ⊗ |Ψ0ancilla i)
which contains only the state |1i1 |1i2 and set γ2 = tanh2 (θ2 ).
(iii) Find the projection

(1 h1|2 ⊗ h1| ⊗ I)|Ψout i.

(iv) Assume that we want to transform the input state (1) into

|Ψout i = α|0i3 + β|1i3 − γ|2i3 .

What is the relation between the coefficients γ1 and γ2 ?

Solution 1. (i) Since

[b†2 b†3 , b2 b3 ] = −b†2 b2 − b†3 b3 − I

and
[b†2 b2 , b†2 b†3 ] = b†2 b†3 , [b†3 b3 , b†2 b†3 ] = b†2 b†3
[b†2 b2 , b2 b3 ] = −b2 b3 , [b†3 b3 , b2 b3 ] = −b2 b3
we have a Lie algebra with the basis b†2 b†3 , b2 b3 , b†2 b2 , b†3 b3 , I. Thus we can
disentangle the operator U23 (θ1 ) as
√ √
U23 (θ1 ) = exp( γ1 b†2 b†3 ) exp(ln(1 − γ1 )(b†2 b2 + b†3 b3 + I)/2) exp(− γ1 b2 b3 ).
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 491

Continuous Variable Teleportation 491

Now we have
√
exp(− γ1 b2 b3 )|0i2 |0i3 = (|0i2 ⊗ |0i3 )
exp(ln(1 − γ1 )(b†2 b2 + b†3 b3 + I)/2)|0i2 |0i3 = exp(ln((1 − γ1 )1/2 )(|0i2 ⊗ |0i3 )
p
= 1 − γ1 (|0i2 ⊗ |0i3 )

and
∞
√
exp( γ1 b†2 b†3 )|0i2 |0i3 =
n/2
X
γ1 |ni2 |ni3 .
n=0

Thus
∞
n/2
p X
U23 (θ1 )|Ψancilla i = 1 − γ1 γ1 (|ni2 ⊗ |ni3 ).
n=0

(ii) Using the result from (i) since the structure of the operator U12 is the
same as U23 we obtain
p √ √
|Ψout i = (1 − γ1 )(1 − γ2 )|1i1 |1i2 ( γ2 α|0i3 + γ1 (1 − 2γ2 )β|1i3
√
+γ1 γ2 (3γ2 − 2)γ|2i3 + Ψothers

(iii) The projection yields the state

p √ √
|Ψout i = (1 − γ1 )(1 − γ2 )( γ2 α|0i3 + γ1 (1 − 2γ2 )β|1i3
√
+γ1 γ2 (3γ2 − 2)γ|2i3 .

(iv) The relation for the parameters is

√ √ √
γ2 = γ1 (1 − 2γ2 ) = −γ1 γ2 (3γ2 − 2)

with the solution

√ √
21 − 7 2 3− 2
γ1 = √ ≈ 0.757, γ2 = ≈ 0.226.
9+4 2 7

Problem 2. Let |βi be a coherent state. Let b and b† be Bose annihilation

and creation operators, respectively. Let D(µ) be the displacement operator
(µ ∈ C). Consider the product state
Z
1
|ψi := d2 β|βi ⊗ |β ∗ i.
π C

This is a maximally entangled continuous-variable state. The state is not

normalized. For teleportation we assume the unknown state |φi to be in
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 492

492 Problems and Solutions

mode 1, the sender’s part of the quantum channel to be in mode 2, and the
receiver’s part in mode 3. Calculate

(12 hψ| ⊗ I3 )(D1† (µ) ⊗ I2 ⊗ I3 )(|φi1 ⊗ |ψi23 )

where I2 is the identity operator acting on mode 2, I3 the identity operator

acting on mode 3, and D1† indicates that the operator acts on mode 1.

Solution 2. Using the completeness relation of coherent states

Z
1
d2 β|βihβ| = I
π C

yields Z
1
|γi = d2 β|βihβ|γi
π C

where we used I|βi = |βi. Applying this expansion and the identity

hγ|βi = hβ ∗ |γ ∗ i

we find
(12 hψ| ⊗ I3 )(D1† (µ) ⊗ I2 ⊗ I3 )(|φi1 ⊗ |ψi23 )

Z Z
1
= d2 βd2 γhγ|D† (µ)|φihγ ∗ |βi|β ∗ i3
π2 C C
Z Z
1
= 2 d2 βd2 γhβ ∗ |γihγ|D† (µ)|φi|β ∗ i3
π C C
Z
1
= d2 βhβ ∗ |D† (µ)|φi|β ∗ i3
π C
Z
1
= d2 β|β ∗ ihβ ∗ |D3† (µ)|φi3
π C
= D3† (µ)|φi3

where we used the identity

Z
1
d2 γhβ ∗ |γihγ|D† (µ)|φi = hβ ∗ |D† (µ)|φi.
π C

We conclude that after the joint measurement, the sender’s state is pro-
jected onto the state which is a unitarily transformed unknown state. Upon
receiving the measurement outcome µ, the receiver recovers the unknown
state by using the appropriate unitary transformation D(µ).
January 30, 2018 15:21 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 493

Continuous Variable Teleportation 493

23.3 Supplementary Problem

Problem 1. Let
∞
−|β|2 /2
X βn
|βi = e √ |ni, β∈C
n=0 n!
be a coherent state. Consider the “coherent Bell states”given by
1
|CΦ+ i = p (|βi ⊗ |βi + | − βi ⊗ | − βi)
N+
1
|CΦ− i = p (|βi ⊗ |βi − | − βi ⊗ | − βi)
N−
1
|CΨ+ i = p (|βi ⊗ | − βi + | − βi ⊗ |βi)
N+
1
|CΨ− i = p (|βi ⊗ | − βi − | − βi ⊗ |βi)
N−
in analogy to the Bell states in C4
1
|Φ+ i = √ (|0i ⊗ |0i + |1i ⊗ |1i)
2
1
|Φ− i = √ (|0i ⊗ |0i − |1i ⊗ |1i)
2
1
|Ψ+ i = √ (|0i ⊗ |1i + |1i ⊗ |0i)
2
− 1
|Ψ i = √ (|0i ⊗ |1i − |1i ⊗ |0i).
2
Note that
N± = 2(1 ± e−4ββ ).
For the Bell states in C4 we have
hΦ+ |Φ− i = 0
etc.. Do we have
hCΦ+ |CΦ− i = 0
etc.? Consider the normalized qubit state
|ψi = a0 |0i + a1 |1i, |a0 |2 + |a1 |2 = 1.
We define the linear operators
R1 (b+ |βi + b− | − βi) = b− |βi + b+ | − βi
R3 (b+ |βi + b− | − βi) = b+ |βi − b− | − βi
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 494

494 Problems and Solutions

in analogy to the Pauli spin matrices σ1 and σ3 . Note that σ1 σ3 = −iσ2 .

Consider the normalized state
1
|γi = p (b+ |βi + b− | − βi)
N+,−

Describe teleportation utilizing the expression


1 1
|ψi ⊗ |CΦ+ i = p √ (|0i ⊗ |βi + |1i ⊗ | − βi) ⊗ |γi
N+ 2
1
+ √ (|0i ⊗ |βi − |1i ⊗ | − βi) ⊗ R3 |γi
2
1
+ √ (|0i ⊗ | − βi + |1i ⊗ |βi) ⊗ R1 |γi
2

1
+ √ (|0i ⊗ | − βi − |1i ⊗ |βi) ⊗ (R1 R3 )|γi .
2
Can this scheme also be applied if we replace the coherent state by squeezed
states?
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 495

Chapter 24

Swapping and Cloning

24.1 Introduction
Swapping and cloning need not only be studied for finite-dimensional sys-
tems but also for continuous variables. We can therefore investigate whether
coherent states |βi and squeezed states |ζi can be swapped or cloned.

24.2 Solved Problems

Problem 1. Can two coherent states be swapped, i.e. can we find a
unitary transformation (swap operator) such that

Uswap (|β1 i ⊗ |β2 i) = |β2 i ⊗ |β1 i

holds? Consider the unitary operator

† ∗
b1 b†2
U (z) := ezb1 b2 −z , z ∈ C.

Solution 1. Yes, we can find a swap operator. From the unitary operator
given above we find U (z)(|0i ⊗ |0i) = |0i ⊗ |0i. Now we have

U (z)(|β1 i ⊗ |β2 i) = U (z)(D(β1 ) ⊗ D(β2 ))|0i ⊗ |0i

= U (z)D1 (β1 )D2 (β2 )U −1 (z)|0i ⊗ |0i

495
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 496

496 Problems and Solutions

where D(β) is the displacement operator. Thus

U (z)D1 (β1 )D2 (β2 )U −1 (z) = U (z) exp(β1 b†1 − β1∗ b1 + β2 b†2 − β2∗ b2 )U (z)−1

and therefore

U (z)D1 (β1 )D2 (β2 )U −1 (z) = exp(β1 (U (z)b1 U (z)−1 )† − β1∗ (U (z)b1 U (z)−1 )
+β2 (U (z)b2 U (z)−1 )† − β2∗ (U (z)b2 U (z)−1 ))
≡ exp(X).

Calculating the unitary transformations in the exponent, we find

z ∗ sin(|z|) ∗
   
z sin(|z|) † ∗
X = cos(|z|)β1 + β2 b1 − cos(|z|)β1 + β 2 b1
|z| |z|
z ∗ sin(|z|)
   
† ∗ z sin(|z|) ∗
+ cos(|z|)β2 − β1 b2 − cos(|z|)β2 − β 1 b2 .
|z| |z|

Thus
z ∗ sin(|z|)
   
z sin(|z|)
exp(X) = D1 cos(|z|)β1 + β2 D2 cos(|z|)β2 − β1
|z| |z|
z ∗ sin(|z|)
   
z sin(|z|)
= D cos(|z|)β1 + β2 ⊗ D cos(|z|)β2 − β1 .
|z| |z|

Therefore, we have

z sin(|z|) z ∗ sin(|z|)
|β1 i ⊗ |β2 i → | cos(|z|)β1 + β2 i ⊗ | cos(|z|)β2 − β1 i.
|z| |z|

If we write z = |z|eiδ , then we can write

|β1 i ⊗ |β2 i → | cos(|z|)β1 + eiδ sin(|z|)β2 i ⊗ | cos(|z|)β2 − e−iδ sin(|z|)β1 i.

Choosing sin(|z|) = 1 yields

|β1 i ⊗ |β2 i → |eiδ β2 i ⊗ |−e−iδ β1 i = |eiδ β2 i ⊗ |e−i(δ+π) β1 i. (1)

Applying the unitary operator

† † † †
V = e−iδb1 b1 ei(δ+π)b2 b2 = e−iδb b
⊗ ei(δ+π)b b

from the left, we find |β1 i ⊗ |β2 i → |β2 i ⊗ |β1 i. If we set β1 = β and β2 = 0
in (1) we obtain

|βi⊗|0i = | cos(|z|)βi⊗|−e−iδ sin(|z|)βi = | cos(|z|)βi⊗|e−i(δ+π) sin(|z|)βi.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 497

Swapping and Cloning 497

Problem 2. We cannot clone coherent states, i.e., we cannot find a

unitary operator which maps

|βi ⊗ |0i → |βi ⊗ |βi.

Use the result from problem 1 (equation (1))

|βi ⊗ |0i → | cos(|z|)βi ⊗ |e−i(δ+π) sin(|z|)βi (1)

to find an approximation.

†
Solution 2. Applying the operator I ⊗ ei(δ+π)b b
to the right-hand side
of (1), we obtain
| cos(|z|)βi ⊗ | sin(|z|)βi.
If we set |z| = π/4 we obtain
   
 β  β
|βi ⊗ |0i →  √ ⊗  √ .
2 2

This is called imperfect cloning.

Problem 3. We consider three infinite-dimensional Hilbert spaces H1 ,

H2 , H3 and the product Hilbert space H3 ⊗ H1 ⊗ H2 with H1 = H2 = H3 .
Consider the heterodyne-current operator Z := b1 + b†2 , where the Bose
annihilation operator acts in the Hilbert space H1 and the Bose creation
operator b†2 acts on the Hilbert space H2 . We have [Z, Z † ] = 0 and the
eigenvalue equation Z|zii12 = z|zii12 with z ∈ C. The eigenstates |zii are
given by
|zii12 = D1 (z)|0ii12 = D2 (z ∗ )|0ii12
where D1 denotes the displacement operator for mode 1, D2 the displace-
ment operator for mode 2 and
∞
1 X
|0ii12 := √ (−1)n |ni1 ⊗ |ni2
π n=0

in the Fock basis (number basis). The expression

1
32 hhz|z
0
ii12 = D1 (z 0 )T13 D3† (z)
π
where
∞
X ∞
X
T13 := |ni13 hn| ≡ (|ni ⊗ I ⊗ I)(I ⊗ I ⊗ hn|)
n=0 n=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 498

498 Problems and Solutions

denotes the transfer operator which obviously satisfies T13 |ψi3 = |ψi1 for
any vector |ψi. For a cloning operation consider the input state in the
product Hilbert space H3 ⊗ H1 ⊗ H2
Z
|ψi = |φi3 ⊗ d2 zf (z, z ∗ )|zii12 .
C

where |φi3 is the original state in the Hilbert space H3 to be cloned in H3

itself and H1 and H2 is an ancillary Hilbert space. The cloning transfor-
mation is realized by the unitary operator
    
1 1 1 1
U = exp (b3 + b†3 ) + (b3 − b†3 ) Z † − (b3 + b†3 ) − (b3 − b†3 ) Z
2 2 2 2

where Z = b1 + b†2 = b ⊗ I + I ⊗ b† . Let |ψiout = U |ψi.

(i) Calculate the commutator [b3 b†1 + b3 b2 , b†3 b1 + b†3 b†2 ] and discuss.
(ii) Evaluate the one-site restricted density matrix ρ3 corresponding to the
state |ψiout for the Hilbert space H3 .
(iii) Evaluate the one-site restricted density matrix ρ1 corresponding to the
state |ψiout for the Hilbert space H1 .
(iv) Compare the two density matrices.

Solution 3. (i) Using [bj , b†k ] = δjk I we find

[b3 b†1 + b3 b2 , b†3 b1 + b†3 b†2 ] = I + b†1 b1 + b†2 b2 + b†1 b†2 + b1 b2 .

The right-hand side does not depend on b3 and b†3 .

(ii) Let |wii12 be an eigenstate of the operator Z. We have ρ = |ψihψ|.
Thus for the partial trace we have to calculate
Z Z Z
ρ3 = d2 w d2 z d2 z 0 f (z, z ∗ )f ∗ (z 0 , z 0∗ )A
C C C

where A ≡ 12 hhw|D3† (z)|φi33 hφ|D3 (z 0 ) ⊗ |zii1212 hhz 0 |wii12 . Using the com-
pleteness and orthogonality of the eigenstates |wii12 of the operator Z we
find Z
ρ3 = d2 z|f (z, z ∗ )|2 D3† (z)|φi33 hφ|D3 (z).
C

(iii) For ρ1 we have to calculate

d2 z d2 z 0
Z Z Z
ρ1 = 2
d w f (z, z ∗ )f ∗ (z 0 , z 0∗ )
C C π C π
×D1 (z)T13 (D3† (w)D3† (z)|φi33 hφ|D3 (z 0 )D3 (w))T31 D3† (z 0 ) .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 499

Swapping and Cloning 499

Using the completeness and orthogonality of the eigenstates |wii of the

operator Z we find after integration over z and z 0
Z
ρ1 = d2 w|fe(w, w∗ )|2 D1† (w)|φi11 hφ|D1 (z)
C

where fe(w, w∗ ) denotes the Fourier transform over the complex plane
Z
1 ∗ ∗
fe(w, w∗ ) = d2 zewz −w z f (z, z ∗ ).
π C

(iv) For
r
2 −|z|2
f (z, z ∗ ) = e
π
one has two identical clones, i.e., ρ3 = ρ1 which are given by the original
state |ψi degraded by Gaussian noise.

Problem 4. Let α ∈ R. Find

†
⊗b−b⊗b† )
eα(b (|1i ⊗ |0i).

Solution 4. We have

(b† ⊗ b − b ⊗ b† )(|1i ⊗ |0i) = −|0i ⊗ |1i

(b† ⊗ b − b ⊗ b† )(−|0i ⊗ |1i) = −|1i ⊗ |0i
(b† ⊗ b − b ⊗ b† )(−|1i ⊗ |0i) = |0i ⊗ |1i
(b† ⊗ b − b ⊗ b† )(|0i ⊗ |1i) = |1i ⊗ |0i.

Therefore we find

α2 α4 α3
   
α(b† ⊗b−b⊗b† )
e = |1i ⊗ |0i 1 − + − · · · + |0i ⊗ |1i −α + − ···
2! 4! 3!
= cos(α)|1i ⊗ |0i − sin(α)|0i ⊗ |1i.

If α = π/2, then

exp(α(b† ⊗ b − b ⊗ b† ))(|1i ⊗ |0i) = −|0i ⊗ |1i

and if α = 3π/2, then

exp(α(b† ⊗ b − b ⊗ b† ))(|1i ⊗ |1i) = |0i ⊗ |1i.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 500

500 Problems and Solutions

24.3 Supplementary Problems

Problem 1. Consider the four Bell states of entangled coherent states

1
|CΦ+ i = p (|βi ⊗ |βi + | − βi ⊗ | − βi)
N+
1
|CΦ− i = p (|βi ⊗ |βi − | − βi ⊗ | − βi)
N+
1
|CΨ+ i = p (|βi ⊗ | − βi + | − βi ⊗ |βi)
N+
1
|CΨ− i = p (|βi ⊗ | − βi − | − βi ⊗ |βi).
N+

Let UBS be the unitary operator of the beam splitter

π 
UBS = exp (b† ⊗ b − b ⊗ b† ) .
4
Find the states

UBS |CΦ+ i, UBS |CΦ− i, UBS |CΨ+ i, UBS |CΨ− i.

Study also the case with the coherent state |β replaced by a squeezed state
|ζi.

Problem 2. (i) Find a unitary operator U such that

U |βi = | − βi.

(ii) Find a unitary operator V such that

V |ζi = | − ζi.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 501

Chapter 25

Homodyne Detection

25.1 Introduction
In optical homodyne detection one mixes a local oscillator field (for example
a coherent state or squeezed state of light) with a signal field at a balanced
(50/50) beam splitter (unitary operator)

UBS = exp(iπ(b†S ⊗ bLO + bS ⊗ b†LO )/4)

where the signal field mode is represented by the Bose operator bS and the
local oscillator is represented by the Bose operator bLO . Then
   †   
bS ⊗ I † UBS (bS ⊗ I)UBS 1 bS ⊗ I − iI ⊗ bLO
UBS UBS = † =√ .
I ⊗ bLO UBS (I ⊗ bLO UBS 2 I ⊗ bLO − ibS ⊗ I

Thus the signal wave is overlapped on a beam splitter with a relatively

strong local oscillator wave in the matching optical mode. The two fields
emerging from the beam splitter are incident on two high-efficiency pho-
todiodes whose output photocurrents are subtracted. The photocurrent
difference is proportional to the value of the electric field operator. Thus
optical homodyne detection corresponds to the difference photon counting
of the two output electric fields from the beam splitter.

501
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 502

502 Problems and Solutions

In heterodyne detection the signal electric field and another electric field,
the auxiliary electric field, feed the same port of a beam splitter. The
local electric field oscillator enters the other port of the beam splitter. The
frequencies of the signal, auxiliary and local oscillator fields are different.
One has ωS + ωA = 2ωL , ωS − ωA = 2ω with ωL  ω.

25.2 Solved Problems

Problem 1. A homodyne detector is constructed by placing a photo-
counter in each arm after the beam splitter and then considering the dif-
ference photocurrent between the two modes

D̂ := eb†1eb1 − eb†2eb2 .

Express the homodyne photocurrent in terms of the input modes b1 , b2 .

Solution 1. Straightforward calculation yields

D̂ = (2τ − 1)(b†1 b1 − b†2 b2 ) + 2 τ (1 − τ )(b†1 b2 + b1 b†2 ).

p

This expression reduces to

D̂ = b†1 b2 + b1 b†2 ≡ b† ⊗ b + b ⊗ b†

for a balanced (τ = 1/2) beam splitter.

Problem 2. Let b1 , b2 be Bose annihilation operators. Consider the

operators
1
d± = √ (b1 ± b2 eiθ )
2
where θ is a phase shift. Let I := d†+ d+ − d†− d− . Find I in terms of the
original operators b1 , b2 .

Solution 2. We have

I = d†+ d+ − d†− d−
1 1
= (b†1 + b†2 e−iθ )(b1 + b2 eiθ ) − (b†1 − b†2 e−iθ )(b1 − b2 eiθ )
2 2
= b†2 b1 e−iθ + b†1 b2 eiθ .

This plays a role in homodyne measurement, where b1 describes the signal

field and b2 describes the local oscillator field.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 503

Homodyne Detection 503

Problem 3. Let bS , bLO , bP be the Bose annihilation operators cor-

responding to the signal, the local oscillator and the photodetector input
field, respectively. Let 0 <  < 1. By shifting the bP field with a constant
phase, we have √ √
bP = bS + i 1 − bLO .
Let
bS = x̂S + iŷS , bLO = x̂LO + iŷLO , bP = x̂P + iŷP
where x̂, ŷ are the (hermitian) quadrature operators. Let
n̂S = b†S bS , n̂LO = b†LO bLO , n̂P = b†P bP
be the photon number operators and let h i be the average with respect to
a quantum state. Find the mean photodetector output hn̂P i.

Solution 3. We obtain
hn̂P i = n̂ + (1 − )n̂ − 2((1 − ))1/2 hx̂S ŷLO − x̂LO ŷS i.

Problem 4. Suppose that b†1 , b†2 are Bose creation operators and b1 , b2
are Bose annihilation operators and I is the identity operator. Consider
the linear operator
Z := b ⊗ I + I ⊗ b†
where b1 := b ⊗ I and b†2 := I ⊗ b† . Thus Z = b1 + b†2 . The operator is
called the heterodyne-current operator. One also finds the notation
Z = bS + bI
where the subscripts S and I, respectively denote the signal mode at fre-
quency ω0 + ∆ω and the imaging mode at frequency ω0 − ∆ω (∆ω  ω).
(i) Calculate the commutator [Z, Z † ].
(ii) Find the states Z(|0i ⊗ |0i), Z † (|0i ⊗ |0i).
(iii) Find the state Z 2 (|0i ⊗ |0i).

Solution 4. (i) We have

Z † = b†1 + b2 ≡ b† ⊗ I + I ⊗ b
and
[Z, Z † ] = (b ⊗ I + I ⊗ b† )(b† ⊗ I + I ⊗ b)
−(b† ⊗ I + I ⊗ b)(b ⊗ I + I ⊗ b† )
= bb† ⊗ I + b ⊗ b + b† ⊗ b† + I ⊗ b† b
−b† b ⊗ I − b† ⊗ b† − b ⊗ b − I ⊗ bb†
= (bb† − b† b) ⊗ I + I ⊗ (b† b − bb† )
= I ⊗ I − I ⊗ I = 0.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 504

504 Problems and Solutions

(ii) We have

Z(|0i ⊗ |0i) = (b ⊗ I + I ⊗ b† )|0i ⊗ |0i

= (b ⊗ I)(|0i ⊗ |0i) + (I ⊗ b† )(|0i ⊗ |0i)
= |0i ⊗ b† |0i = |0i ⊗ |1i

since b|0i = 0. Analogously

Z † |0i ⊗ |0i = b† |0i ⊗ |0i = |1i ⊗ |0i.

(iii) We find √
Z 2 (|0i ⊗ |0i) = 2(|0i ⊗ |2i).

25.3 Supplementary Problems

Problem 1. (i) Show that the spectrum of the operator

K̂ = b†1 b2 + b1 b†2 ≡ b† ⊗ b + b ⊗ b†

is discrete and coincides with the set Z of relative integers. Note that

K̂(|0i ⊗ |0i) = 0(|0i ⊗ |0i)

√ √
K̂(|1ii ⊗ |1i) = 2|2i ⊗ |0i + 2|0i ⊗ |2i.
(ii) Find

(hβ| ⊗ hβ|)K̂(|βi ⊗ |βi), (hζ| ⊗ hζ|)K̂(|ζi ⊗ |ζi).

Problem 2. Let b1 , b2 be Bose annihilation operators and

K̂ = b1 cos(φ) + b2 sin(φ).

Show that

K̂ † K̂ = b†1 b1 cos2 (φ) + b†2 b2 sin2 (φ) + (b†1 b2 + b1 b†2 ) sin(φ) cos(φ).

Study the case that b1 , b2 are Fermi annihilation operators.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 505

Chapter 26

Hamilton Operators

26.1 Introduction
Most experimental realizations of quantum logic gates (Hadamard gate,
quantum phase gate, controlled-NOT gate) involve several qubits and num-
ber states. A Hamilton operator Ĥ must describe the interaction. Thus in
quantum computing we are faced with two problems. One is to determine
the Hamilton operator Ĥ for the system such that the time-evolution

exp(−iĤt/~)

represents the execution of the computation. The other one is to build the
hardware described by this Hamilton operator.

For example the Hamilton operator that produces squeezed states is given
by
Ĥ = ~ω0 b† b + ~κ(b2 eiωt + (b† )2 e−iωt )

where b† , b are Bose creation and annihilation operators, ω0 is the fre-

quency of the degenerate signal/idler mode,  is the classical pump field of
frequency ω and κ is the coupling constant between the pump and signal
modes.

505
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 506

506 Problems and Solutions

26.2 Solved Problems

Problem 1. Consider a single continuous variable corresponding to a
linear operator X. Let P be the operator of the conjugate variable, i.e.

[X, P ] = iI. (1)

Consider the Kerr-Hamilton operator

K = H 2 = (X 2 + P 2 )2 .

The Kerr-Hamilton operator corresponds to a χ3 process in nonlinear op-

tics. The linear unbounded operators X and P could correspond to quadra-
ture amplitudes of a mode of the electromagnetic field. The quadrature
amplitudes are the real and imaginary parts of the complex electric field.
Let
1
S := (XP + P X).
2
Calculate the commutators

[K, X], [K, P ], [X, [K, S]], [P, [K, S]].

Discuss.

Solution 1. Since

P 2 X 2 = X 2 P 2 − 4iXP − 2I, P 2 X = XP 2 − 2iP

we find
i
[K, X] = (X 2 P + P X 2 + 2P 3 )
2
i
[K, P ] = − (P 2 X + XP 2 + 2X 3 )
2
[X, [K, S]] = P 3
[P, [K, S]] = X 3 .

Thus the algebra generated by X, P , H, S, K by calculating commutators

includes all third order polynomials in X and P . We can construct Hamil-
ton operators that are arbitrary hermitian polynomials in any order of X
and P . We have

[P 3 , P m X n ] = iP m+2 X n−1 + lower order terms

and
[X 3 , P m X n ] = iP m−1 X n+2 + lower order terms.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 507

Hamilton Operators 507

Problem 2. A nonlinear four-wave mixing device can be described by

the Hamilton operator

1 1
Ĥ = ~ω(b†1 b2 + b1 b†2 )2 ≡ ~ω(b† ⊗ b + b ⊗ b† )2 .
4 4
(i) Let |ni be a number state. Consider the normalized input state |ni⊗|0i.
Find the normalized state

exp(−iĤt/~)(|ni ⊗ |0i)

for the interaction time t = π/ω.

(ii) Consider the normalized input state |ni ⊗ |0i. Find the state

exp(−iĤt/~)(|ni ⊗ |0i)

for the interaction time t = π/(2ω).

Solution 2. (i) For the interaction time t = π/ω the output state is
i|0i ⊗ |ni (for n even) and exp(−iπ/4)|ni ⊗ |0i for n odd. Thus the device
acts as an even-odd filter, switching the even numbers from one mode to
the other. Under these operating conditions it can be used as a device
to measure parity without counting the photon number. It is sufficient to
detect any photons in either of the output channels.
(ii) If the interaction time is t = π/(2ω) the output state will have the form

1
√ (|ni ⊗ |0i + e−i(n+1)π/2 |0i ⊗ |ni)
2

which is a maximally entangled state for n photons.

Problem 3. Cross phase modulation is described by the Hamilton oper-

ator
Ĥ = −~ωb†1 b1 b†2 b2 = −~ω(b† b ⊗ b† b)
where ω is a function of the third order susceptibility χ(3) . Consider the
two-mode number state |mi ⊗ |ni, i.e. m, n = 0, 1, 2, . . .. Find the state
exp(−iĤt/~)(|mi ⊗ |ni).

Solution 3. Since b1 = b ⊗ I and b2 = I ⊗ b we have

b†1 b1 |mi ⊗ |ni = m|mi ⊗ |ni, b†2 b2 |mi ⊗ |ni = n|mi ⊗ |ni.

Thus
exp(−iĤt/~)|mi ⊗ |ni = eiωtmn |mi ⊗ |ni.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 508

508 Problems and Solutions

Problem 4. A Kerr medium is nonlinear in the sense that its refractive

index n has a component which varies with the intensity of the propagating
electric field E, that is
n = n0 + n2 |E|2
where n0 and n2 are constants. For a single-mode field, described by Bose
creation and annihilation operators b† and b, propagating through a low-loss
Kerr media, the interaction Hamilton operator can be written as

Ĥ = χb†2 b2 .

(i) Show that the number state |ni is an eigenstate.

(ii) Assume that the initial state is a coherent state |βi. Find |β(t)i.
(iii) Let χt = πr/s where r and s are mutually prime with r < s. Write
exp(−iπrn2 /s) as a discrete Fourier transform. Express |β(t)i using this
expansion.

Solution 4. (i) Since b†2 b2 ≡ b† b(b† b − I) and b† b|ni = n|ni, we have

Ĥ|ni = χ(n2 − n)|ni.

Thus the eigenvalues are χ(n2 − n).

(ii) The solution of the time dependent Schrödinger equation (~ = 1)

d|βi
i = Ĥ|βi
dt
is given by
|β(t)i = exp(−iĤt)|βi.
Using the result from (i) we find
∞
X 2
|β(t)i = cn e−iχt(n −n)
|ni
n=0

where
βn
cn := exp(−|β|2 /2) √ .
n!
Since n2 − n is always an even number, the system will revive whenever χt
is a multiple of π.
(iii) Let χt = πr/s where r, s are mutually prime with r < s. Then we can
write the quadratic (in n) phase in terms of linear phases using the discrete
Fourier transform
`−1
X
exp(−iπn2 r/s) = a(r,s)
p exp(−2πipn/`)
p=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 509

Hamilton Operators 509

where 
s if r is odd, s is even or vice-versa
`=
2s if both r and s are odd.
Thus
`−1
1X
a(r,s)
p = exp(−iπrk 2 /s + 2πipk/`)
`
k=0
and
`−1
X
|β(t)i = a(r,s)
p |β exp(iπ(r/s − 2p/`))i.
p=0

Problem 5. The Hamilton operator for the second harmonic generation

can be written as
κ
Ĥ = i~ (b†2 bsh − b2 b†sh )
2
where b is the fundamental cavity mode Bose operator, bsh is the second-
harmonic mode Bose operator and κ is the nonlinear coupling. Using the
Heisenberg equation of motion find the time evolution of b and bsh .

Solution 5. The Heisenberg equation of motion of an operator  is given

by
dÂ
i~
= [Â, Ĥ](t).
dt
The commutation relations are given by

[b, b† ] = I, [bsh , b†sh ] = I

[b, b] = [bsh , bsh ] = [b, bsh ] = [b, b†sh ] = 0.

Thus we find the operator-valued nonlinear differential equations
db dbsh κ
= κb† bsh , = − b2 .
dt dt 2
In a more realistic model, cavity photon losses must be taken into account.

Problem 6. A single spin- 12 particle is placed on a cantilever tip. The

tip can oscillate only in the z-direction. A ferromagnetic particle, whose
magnetic moment points in the positive z-direction, produces a non-uniform
magnetic field at the spin. A uniform magnetic field, B0 , oriented in the
positive z-direction, determines the ground state of the spin. A rotating
magnetic field, B1 (t), induces transitions between the ground state and
excited states of the spin. It is given by

Bx (t) = B1 cos(ωt + φ(t)), By (t) = −B1 sin(ωt + φ(t)), Bz (t) = 0

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 510

510 Problems and Solutions

where φ(t) describes a smooth change in phase required for a cyclic adia-
batic inversion of the spin
|dφ(t)/dt|  ω.
In the reference frame rotating with B1 (t), the time-dependent Hamilton
operator is given by
P2 m∗ ω 2 Z 2
 
dφ ∂Bz
Ĥ(t) = z ∗ + c c − ~ ωL − ω − S3 − ~ω1 S1 − gµ ZS3
2mc 2 dt ∂Z
where Z is the coordinate of the oscillator which describes the dynamics of
the quasi-classical cantilever tip, Pz is its momentum, m∗c and ωc are the
effective mass and the frequency of the cantilever, S3 and S1 are the z−
and the x− component of the spin,
   
1 0 1 1 1 0
S1 = , S3 = ,
2 1 0 2 0 −1
ωL is its Larmor frequency, ω1 is the Rabi frequency (the frequency of the
spin precession around the magnetic field B1 (t) at the resonance condition
ω = ωL , dφ/dt = 0), g and µ are the g-factors and the magnetic moment of
the spin and we defined m∗c = mc /4 as the effective cantilever mass. The
operator acts in the product Hilbert space L2 (R) ⊗ C2 . One sets
ωc = (kc /m∗c )1/2 , ωL = γBz , ω1 = γB1
where γ = gµ/~ is the gyromagnetic ratio of the spin, mc and kc are the
mass and the force constant of the cantilever, Bz includes the uniform
magnetic field B0 and the magnetic field produced by the ferromagnetic
particle.
(i) Cast the Hamilton operator in dimensionless form Ĥ/~ωc → K̂ by
introducing the quantities
p p
E0 := ~ωc , F0 := kc E0 , Z0 := E0 /kc , P0 := ~/Z0
with ω = ωL and using the dimensionless time τ := ωc t.
(ii) The dimensionless time-dependent Schrödinger equation
∂Ψ
i = K̂Ψ
∂τ
where  
Ψ1 (τ, z)
Ψ(τ, z) =
Ψ2 (τ, z)
can be solved using the expansions
∞
X ∞
X
Ψ1 (τ, z) = An (τ )|ni, Ψ2 (τ, z) = Bn (τ )|ni
n=0 n=0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 511

Hamilton Operators 511

2
|ni = π 1/4 2n/2 (n!)1/2 e−z /2
Hn (z)
where { |ni : n = 0, 1, . . . } are number states. Here Hn (z) are the Her-
mitian polynomials. Find the time evolution of the complex expansion
coefficients An and Bn .
(iii) What would be an initial state closest to the classical limit?

Solution 6. (i) Since Ĥ/(~ωc ) → K̂ we find

Ĥ 1 dφ
K̂ = = (p2z + z 2 ) + S3 − S1 − 2ηzS3
~ωc 2 dτ
where we used ωL = ω and
Pz Z ω1 gµ ∂Bz
pz := , z := ,  := , η= , ωc dt = dτ.
P0 Z0 ωc 2Fc ∂Z
(ii) Inserting the series expansions into the dimensionless Schrödinger equa-
tion we find the system of linear differential equations with time-dependent
coefficients for the complex amplitudes An (τ ) and Bn (τ )

η √ √
 
dAn 1 1 dφ

i = n+ + An − √ nAn−1 + n + 1An+1 − Bn
dτ 2 2 dτ 2 2

η √ √
 
dBn 1 1 dφ

i = n+ + Bn + √ nBn−1 + n + 1Bn+1 − An
dτ 2 2 dτ 2 2
where we used the Bose operators b and b† defined by
√ √
b|ni = n|n − 1i, b† |ni = n + 1|n + 1i

and  
1 2 1
pz + z 2 |ni =


n+ |ni
2 2
with
1 i
z := √ (b† + b), pz := √ (b† − b), [b, b† ] = I.
2 2
(iii) We can choose the coherent state
∞
X βn
Ψ1 (z, 0) = An (0)|ni, Ψ2 (z, 0) = 0, An (0) = √ exp(−|β|2 /2).
n=0 n!

Problem 7. Consider two Bose-Einstein condensates which both oc-

cupy the ground-state of their respective traps. They are described by the
atom Bose annihilation (creation) operators b1 (b†1 ) and b2 (b†2 ). Atoms are
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 512

512 Problems and Solutions

released from each trap with momenta (wave vectors) k1 and k2 , respec-
tively, producing an interference pattern which enables a relative phase to
be measured. The intensity I(x, t) of the atomic field is given by

I(x, t) = I0 hψ|(b†1 (t)eik1 ·x + b†2 (t)eik2 ·x )(b1 (t)e−ik1 ·x + b2 (t)e−ik2 ·x )|ψi (1)

where I0 is the single atom intensity. Atoms within each condensate collide.
This can be described using the Hamilton operator
1
Ĥ = ~χ((b†1 b1 )2 + (b†2 b2 )2 ) (2)
2
where χ is the collision rate between the atoms within each condensate.
Cross-collisions between the two condensates, described by the term b†1 b1 b†2 b2
could also be included. Using the Hamilton operator given by (2) the
intensityI(x, t) is given by

I(x, t) = I0 (hψ|b†1 b1 |ψi + hψ|b†2 b2 |ψi + hψ|b†1 exp(iχt(b†1 b1 − b†2 b2 ))b2 |ψie−iφ(x)
+hψ|b†2 exp(−iχt(b†1 b1 − b†2 b2 ))b1 |ψieiφ(x) )

where φ(x) := (k2 − k1 ) · x. Calculate I(x, t) for the product state

|ψi = |β1 i ⊗ |β2 i

where |β1 i and |β2 i are coherent states.

Solution 7. Since b|βi = β|βi, hβ|b† = hβ|β ∗ , and

∞ ∞
† X 1 iχt X 1 iχt
eiχtb b |βi = (e − 1)(b† )j bj |βi = (e − 1)(b† )j β j
j=0
j! j=0
j!

we have
†
hβ|eiχtb b |βi = exp((eiχt − 1)β ∗ β).
We also have
† †
hβ|eiχtb b b|βi = β exp((eiχt −1)β ∗ β), hβ|b† eiχtb b |βi = β ∗ exp((eiχt −1)β ∗ β).

Since b†1 b1 = b† b ⊗ I and b†2 b2 = I ⊗ b† b we have

† † † † †
b⊗I) −iχt(I⊗b† b)
eiχt(b1 b1 −b2 b2 ) = eiχtb1 b1 e−iχtb2 b2 = eiχt(b e
iχtb† b −iχtb† b † †
= (e ⊗ I)(I ⊗ e ) = eiχtb b
⊗ e−iχtb b .

Thus
hψ|b†1 b1 |ψi = β1∗ β1 , hψ|b†2 b2 |ψi = β2∗ β2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 513

Hamilton Operators 513

and
† †
hψ|b†1 eiχt(b1 b1 −b2 b2 ) b2 |ψie−iφ(x) = β1∗ β2 exp((eiχt − 1)β1∗ β1 )
× exp((e−iχt − 1)β2∗ β2 )e−iφ(x)

† †
hψ|b†2 e−iχt(b1 b1 −b2 b2 ) b1 |ψieiφ(x) = β1 β2∗ exp((e−iχt − 1)β1∗ β1 )
× exp((eiχt − 1)β2∗ β2 )eiφ(x) .

Problem 8. Consider the Hamilton operator Ĥ of a coupled one Bose

one Fermi system

Ĥ = ωb† b ⊗ IF + JIB ⊗ c† c + α(b† ⊗ c + b ⊗ c† )

where ω, J, α are positive quantities, IB is the identity operator in the

space of the Bose operators and IF is the identity operator in the space
of the Fermi operators. Here b and b† are Bose annihilation and creation
operators, respectively, and c and c† are Fermi annihilation and creation
operators, respectively. The commutation and anticommutation relations
are
[b, b† ] = IB , [b, b] = [b† , b† ] = 0
[c, c† ]+ = IF , [c, c]+ = [c† , c† ]+ = 0.
Find the eigenvalues of the Hamilton operator Ĥ. Find the eigenvectors.
Use the matrix representations for c† , c, b† , b, i.e.
   
† 0 0 0 1
c = , c= .
1 0 0 0

For the Fermi states we set

 
1
|0iF = .
0

Thus  
† 0
|1iF = c |0iF = .
1
For the Bose operators we have

0 0 0 0 ... 0 1 √0 0 0 ...
   
1 √0 0 0 ...  0 0 2 √0 0 ...
b† =  0 2 √0 0 ... , b = 0 0 0 3 √0 ....
   
0 0 3 0 ... 0 0 0 0 4 ...
   
... ...
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 514

514 Problems and Solutions

For the number states of the Bose operators we have

1
|niB = √ b†n |0iB
n!
where n = 0, 1, 2, . . ..

Solution 8. Using the matrix representation for the number operators

c† c and b† b given above we find
 
† 0 0
c c=
0 1

and
b† b = diag(0, 1, 2, . . .).
Thus it follows that
 
1 0
b† b ⊗ IF = diag(0, 1, 2, . . .) ⊗ = diag(0, 0, 1, 1, 2, 2, . . .)
0 1

and
 
† 0 0
IB ⊗ c c = diag(1, 1, 1, . . .) ⊗ = diag(0, 1, 0, 1, . . .).
0 1

For the interacting terms we find

   √   √ 
0 1 0 2 0 3
b ⊗ c† = (0) ⊕ ⊕ ⊕ ⊕ ···
0 0 0 0 0 0
     
0 0 √0 0 √0 0
b† ⊗ c = (0) ⊕ ⊕ ⊕ ⊕ ···
1 0 2 0 3 0
where ⊕ denotes the direct sum of matrices. Adding up the matrices we
obtain the matrix representation of the Hamilton operator
   √   √ 
J α ω+J 2α 2ω√+ J 3α
Ĥ = (0) ⊕ ⊕ √ ⊕ ⊕ ···
α ω 2α 2ω 3α 3ω
 √ 
nω + J n + 1α
⊕ √ ⊕ ··· .
n + 1α (n + 1)ω

Thus one eigenvalue is 0. The eigenvector for the eigenvalue 0 is given by

|0iB ⊗ |0iF . Thus this state is not entangled. To find the other eigenvalues
we have to find the eigenvalues of the 2 × 2 matrices
 √ 
√nω + J n + 1α
n + 1α (n + 1)ω
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 515

Hamilton Operators 515

where n = 0, 1, 2, . . .. We find
r
J ω 1
E± (n) = + + nω ± (J − ω)2 + (n + 1)α2 .
2 2 4
Whether 0 is the lowest eigenvalue depends on the values of J, ω and α.
Consider the case with n = 0. Then
r
J ω 1
E− (n = 0) = + − (J − ω)2 + α2 .
2 2 4
The eigenvector for the eigenvalue E− (n = 0) is given by (α 6= 0)
r !
1 1 ω J
|0iB ⊗ |1iF + − (J − ω)2 + α2 + − |1iB ⊗ |0iF .
α 4 2 2
p
Thus this state is entangled except if − (J − ω)2 /4 + α2 + ω/2 − J/2 = 0.
The condition E− (n = 0) = 0 yields α2 = Jω. Then we find the state
J
|0iB ⊗ |1iF − |1iB ⊗ |0iF .
α
This state is also entangled if J 6= 0. Thus we have an unentangled state
with eigenvalue 0 and for given parameter values of J, ω and α we can have
an entangled state with this eigenvalue.

Problem 9. Let b, b† be Bose annihilation and creation operators. Let

   
1 0 1 1 0 0
σ+ := (σ1 + iσ2 ) = , σ− := (σ1 − iσ2 ) = .
2 0 0 2 1 0
Consider the Hamilton operator
Ĥ = b† ⊗ σ− + b ⊗ σ+
which describes a single atom coupled to a single mode of an electromag-
netic field. σ± act on the atom and b, b† act on the field.
(i) Calculate Ĥ 2 .
(ii) Calculate the commutator [b† ⊗ σ− , b ⊗ σ+ ].
(iii) Let |ni be a number state and |βi be a coherent state. Calculate

A(n, β) := (hn| ⊗ I2 )eiθĤ (|βi ⊗ I2 )

where θ ∈ R and I2 is 2 × 2 unit matrix.

Solution 9. (i) Since σ− σ− = 0, σ+ σ+ = 0 and

   
1 0 0 0
σ+ σ− = , σ− σ+ =
0 0 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 516

516 Problems and Solutions

we find

(b† ⊗ σ− + b ⊗ σ+ )(b† ⊗ σ− + b ⊗ σ+ ) = bb† ⊗ σ+ σ− + b† b ⊗ σ− σ+ .

Thus we can write in matrix notation

 † 
bb 0
.
0 b† b

(ii) We have

[b† ⊗ σ− , b ⊗ σ+ ] = b† b ⊗ σ− σ+ − bb† ⊗ σ+ σ− .
√
(iii) Since b|βi = β|βi and b|ni = n|n − 1i we obtain
√ √ √ !
i n
|β|2
2 cos(θ n) β sin(θ n)
A(n, α) = exp(−|β| ) √ √ .
n! √ in sin(θ n + 1) cos(θ n + 1)
n+1

Problem 10. Consider the model Hamilton operator for ions trapped
inside an optical cavity
Ĥ := Ĥ0 + V̂
where
 
1 ~ω0
Ĥ0 = ~νa† a + Ia ⊗ Ib ⊗ I2 + Ia ⊗ ~ωc b† b ⊗ I2 + Ia ⊗ Ib ⊗ σ3
2 2

and

V̂ = ~Ω(exp(iηL (a† + a) − i(ωL t + φ)Ia ) ⊗ Ib ⊗ σ+ + h.c.)

+~g sin(ηc (a† + a)) ⊗ (b† + b) ⊗ (σ+ + σ− ).

Here a† (a) and b† (b) are Bose creation (annihilation) operators for the
vibrational phonon and the cavity field photon, respectively and ω0 is the
transition frequency of the two-level ion. The ion-phonon and ion-cavity
coupling constants are Ω and g, and σk (k = z, +, −) are the Pauli operators
describing the internal state of the ion. Thus we consider a two-level ion
radiated by the single mode cavity field of frequency ωc and an external
laser field of frequency ωL . The operators Ia , Ib and I2 are the identity
operators in their respective Hilbert spaces, where I2 is the 2 × 2 unit
matrix. Thus we have a tripartite system. The parameters ηL and ηc are
the Lamb-Dicke parameters.
(i) Consider the unitary operator

U0 (t) = exp(−iĤ0 t/~).

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 517

Hamilton Operators 517

Calculate (interaction picture)

HI (t) = U0† (t)V̂ U0 (t) ≡ exp(iĤ0 t/~)V̂ exp(−iĤ0 t/~).

(ii) Discuss how a Hadamard gate can be realized.

Solution 10. (i) Straightforward calculation yields

 
ĤI (t) = ~Ω Ô0L exp(i(δ0L t − φ)) ⊗ Ib ⊗ σ+
∞
!
X
k
+~Ω (iηL ) ÔkL ak exp(i((δ0L − kν)t − φ)) ⊗ Ib ⊗ σ+
k=1
∞
!
X
+~Ω (iηL )k a†k ÔkL exp(i((δ0L + kν)t − φ)) ⊗ Ib ⊗ σ+
k=1
 
∞
X
+~g  (ik−1 ηck )a†k Ôkc exp(i(δ0c + kν + 2ωc )t) ⊗ b† ⊗ σ+
k=1,3,...
 
∞
X
+~g  (ik−1 ηck )Ôkc ak exp(i(δ0c − kν + 2ωc )t) ⊗ b† ⊗ σ+
k=1,3,...
 
∞
X
+~g  (ik−1 ηck )Ôkc ak exp(i(δ0c − kν)t) ⊗ b ⊗ σ+
k=1,3,...
 
∞
X
+~g  (ik−1 ηck )a†k Ôkc exp(i(δ0c + kν)t) ⊗ b ⊗ σ+
k=1,3,...

+h.c.

where
∞ ∞
2
(iηL )2p a†p ap −ηc2 (iηc )2p a†p ap
 X  X
−ηL
ÔkL := exp , Ôkc := exp
2 p=0
p!(p + k)! 2 p=0
p!(p + k)!

and
δ0L := ω0 − ωL , δ0c := ω0 − ωc .
(ii) A basis is
|mi ⊗ |ni ⊗ |gi, |mi ⊗ |ni ⊗ |ei
where m = 0, 1, . . . , ∞ denotes the state of ionic vibrational motion, n =
0, 1, . . . , ∞ denotes the state of the quantized cavity field and |gi and |ei
denote the ground state and excited state, respectively for the two-level
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 518

518 Problems and Solutions

ion. Using appropriate values for the parameters and the time we can find
an implementation of the Hadamard gate

1
|mi ⊗ |0i ⊗ |gi → (|mi ⊗ |0i ⊗ |gi + |mi ⊗ |0i ⊗ |ei)
2
1
|mi ⊗ |0i ⊗ |ei → (|mi ⊗ |0i ⊗ |gi − |mi ⊗ |0i ⊗ |ei).
2

Problem 11. Consider a square lattice with lattice constant a and pe-
riodic boundary conditions. Let c†j (cj ) denotes the creation (annihilation)
operator for an electron in the Wannier state at the lattice site j. The
Hamilton operator Ĥ of spinless tight-binding electrons in the presence of
a magnetic field can be written as

tj1 j2 c†j1 cj2

X
Ĥ =
hj1 ,j2 i

with !
Z j2
e
tj1 j2 = −t exp −i A · dr
~ j1

The summation hj1 , j2 i runs over the nearest neighbour site on the square
lattice. The uniform magnetic field B is applied in z-direction. Choosing
the Landau gauge
A = B(0, x, 0)
the line integral in tj1 ,j2 can be written as
Z j2 
e 0 j1 = (m, n) j2 = (m + 1, n)
A · dr =
~ j1 −2πmΦ/Φ0 j1 = (m, n) j2 = (m, n + 1)

where the integers m and n refer to the x = 1, y = 2 coordinates of the

square lattice sites. Φ = Ba2 is the magnetic flux through a unit plaquette.
Φ0 stands for the magnetic flux quantum h/e. Find the Hamilton operator
in the Bloch representation. The Fourier transform is given by

1 X i(k1 j1 +k2 j2 )
ck = √ e cj
N j

with the inverse

1 X −i(k1 j1 +k2 j2 )
cj = √ e ck
N k
where k runs over the first Brillouin zone.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 519

Hamilton Operators 519

Solution 11. We obtain the Hamilton operator in Bloch representation

X
Ĥ = −2t cos(k1 a)c† (k)c(k)
k
X πΦ
−t exp(−ik2 a)c† (k1 + 2 , k2 )c(k1 , k2 )
aΦ0
k

† πΦ
+ exp(ik2 a)c (k1 − 2 , k2 )c(k1 , k2 )
aΦ0

where c† (k) creates an electron in the Block state with wave vector k and
c(k) annihilates an electron in the Block state with wave vector k. If
Φ/Φ0 = p/q is rational, the magnetic Brillouin zone can be reduced to
0 ≤ k1 ≤ 2π/a and 0 ≤ k2 ≤ 2π/(qa). We obtain
   
πΦ πΦ
c k1 + 2 `, k2 = c k1 + 2 (` + q), k2
aΦ0 aΦ0

with ` are integers.

26.3 Supplementary Problems

Problem 1. Let b†1 , b†2 be Bose creation operators. Consider the Hamilton
operator
Ĥ = ~ωb†1 b1 (b†2 + b2 ).
Find the unitary operator

U = exp(−iĤt/~) = exp(−iωtb†1 b1 (b†2 + b2 )).

Problem 2. Let b†1 , b†2 be Bose creation operators. Study the Hamilton
operator
κ1 † 2 2 κ2 † 2 2
Ĥ = ~ω1 b†1 b1 +~ω2 b†2 b2 + (b ) b + (b ) b +~ω(b†1 b2 +b1 b†2 )+κ12 b†1 b1 b†2 b2 .
2 1 1 2 2 2

Problem 3. Let |βi be a coherent state. Then the operator

U = I − 2|βihβ|
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 520

520 Problems and Solutions

is unitary, i.e. U U † = I. Find a self-adjoint operator K such that

U = exp(−iK).

Study also the cases

V = I − 2|ζihζ|
where |ζi is a squeezed state and

W = I − 2|nihn|

where |ni is a number state.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 521

Bibliography
Books

Barnett S. M. and Radmore P. M.

Methods in Theoretical Quantum Optics, Oxford University Press (1997)

Bell J. S.
Speakable and unspeakable in quantum mechanics, Cambridge University
Press (1989)

Benenti G., Casati G. and Strini G.

Principles of Quantum Computing and Information, Volume I: Basic Con-
cepts, World Scientific (2007)

Blank J., Exner P. and Havliček M.

Hilbert Space Operators in Quantum Physics, second edition, Springer (2008)

Bohm A.
Quantum Mechanics, Third edition, Springer (1994)

Bouwmeister D., Ekert A. K. and Zeilinger A.

The Physics of Quantum Information: Quantum Cryptography. Quantum
Teleportation, Quantum Computing, Springer (2000)

Braunstein S. L. and Pati A. K.

Quantum Information with Continuous Variables, Kluwer (2003)

Cerf N. J., Leuchs G. and Polzik E. S.

Quantum Information with continuous variables of atoms and light, Impe-
rial College Press (2007)

Constantinescu F. and Magyari E.

Problems in Quantum Mechanics, Pergamon Press (1971)

Dirac P. A. M.
The Principle of Quantum Mechanics, Clarendon Press (1958)

Flügge S.
Practical Quantum Mechanics, Springer (1974)

521
January 25, 2018 8:58 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 522

522 Bibliography

Gasiorowicz S.
Quantum Physics, Wiley (1974)

Gilmore R.
Lie Groups, Lie Algebras, and Some of Their Applications, Wiley (1974)

Glimm J. and Jaffe A.

Quantum Physics, Springer (1981)

Gruska J.
Quantum Computing, McGraw-Hill (1999)

Hardy Y. and Steeb W.-H.

Classical and Quantum Computing with C++ and Java Simulations
Birkhauser-Verlag (2002)

Hayashi M.
Quantum Information: An Introduction, Springer (2010)

Hirvensalo M.
Quantum Computing, second edition, Springer (2004)

Kaye P., Laflamme R. and Mosca M.

An Introduction to Quantum Computing, Oxford University Press (2007)

Kim Y. S. and Noz M. E.

Phase Space Picture of Quantum Mechanics, World Scientific (1991)

Kitaev A. Yu., Shen A. H. and Vyalyi M. N.,

Classical and Quantum Computation, American Mathematical Society (2002)

Klauder J. R. and Skagerstam B.

Coherent States - Applications in Physics and Mathematical Physics, World
Scientific (1985)

Kowalski K. and Steeb W.-H.

Nonlinear Dynamical Systems and Carleman Linearization, World Scien-
tific, (1991)

Lewand R. E.
Cryptological Mathematics, Mathematical Association of America (2000)

Loudon R.
The Quantum Theory of Light, (3rd edition), Oxford University Press (2000)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 523

Bibliography 523

Louisell W. H.
Quantum Statistical Properties of Radiation, Wiley (1973)

Mandel L. and Wolf E.

Optical coherence and Quantum Physics, Cambridge University Press (1995)

Mermin N. D.
Quantum Computer Science, Cambridge University Press (2007)

Merzbacher E.
Quantum Mechanics, 2nd edition, Wiley (1970)

Nielsen M. A. and Chuang I. L.

Quantum Computation and Quantum Information, Cambridge University
Press (2000)

Prugovećki E.
Quantum Mechanics in Hilbert Space, second edition, Academic Press (1981)

Rieffel E. and Polak W.

Quantum Computing: A Gentle Introduction, MIT Press 2011

Sakurai J. J.
Modern Quantum Mechanics, Addison-Wesley (1994)

Scully M. O. and Zubairy M. S.

Quantum Optics, Cambridge University Press (1997)

Steeb W.-H. and Hardy Y.

Matrix Calculus and Kronecker Product: A Practical Approach to Linear
and Multilinear Algebra, second edition, World Scientific (2010)

Steeb W.-H.
Continuous Symmetries, Lie Algebras, Differential Equations and Com-
puter Algebra, second edition, World Scientific (2007)

Steeb W.-H.
Hilbert Spaces, Wavelets, Generalized Functions and Quantum Mechanics,
Kluwer Academic Publishers (1998)

Steeb W.-H. and Hardy Y.

Quantum Mechanics using Computer Algebra, second edition, World Sci-
entific (2010)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 524

524 Bibliography

Steeb W.-H.
Problems and Solutions in Theoretical and Mathematical Physics, Third
Edition, Volume I: Introductory Level, World Scientific (2010)

Steeb W.-H.
Problems and Solutions in Theoretical and Mathematical Physics, Third
Edition, Volume II: Advanced Level, World Scientific (2010)

Steeb W.-H. and Hardy Y.

Bose, Spin and Fermi Systems: Problems and Solutions, World Scientific
(2015)

Stolze J. and Suter D.

Quantum Computing: A short course from theory to experiment, Wiley
(2008)

von Neumann J.
Mathematical Foundations of Quantum Mechanics, Princeton University
Press (1955)

Walls D. F. and Milburn G. J.

Quantum Optics, Springer (1994)

Wilde M. M.
Quantum Information Theory, Cambridge University Press (2017)

Papers

Part I

Alber G., Delgado A., Gisin N. and Jex I., Efficient bipartite quantum state
purification in arbitrary dimensional Hilbert spaces, J. Phys. A: Math. Gen.
34, 8821 (2001)

Ban M., Information and Entropy in Quantum Measurement Processes, Int.

J. Theor. Phys. 37, 2491 (1998)

Bennett C. H., DiVincenzo D. P., Smolin J. A. and Wootters W. K., Mixed

State Entanglement and Quantum Error Correction, Phys. Rev. A 54,
3824 (1996)

Bennett C. H., Brassard G., Popescu S., Schumacher B., Smolin J. A. and
Wootters W.K., Purification of Noisy Entanglement and Faithful Telepor-
tation via Noisy Channels, Phys. Rev. Lett. 77, 722 (1996)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 525

Bibliography 525

Benioff P., Models of quantum Turing machines, Fortsch. Phys. 46, 423
(1997)

Berman G. P., Borgonovi F., Chapline G., Gurvitz S. A., Hammel P. C.,
Pelekhov D. V., Suter A. and Tsifrinovich V. I., Application of Magnetic
Resonance Force Microscopy Cyclic Adiabatic Inversion for a Single-Spin
Measurement, J. Phys. A : Math. Gen. 36, 4417 (2003)

Brandt H. E., Riemannian Geometry of Quantum Computation, Nonlinear

Analysis: Theory, Methods and Applications 71, e474 (2009)

Brassard G., Braunstein S. L. and Cleve R. Teleportation as a quantum

computation, Physica D 120, 43 (1998)

Brassard G., Cleve R. and Tapp A., The cost of Exactly Simulating Quan-
tum Entanglement with Classical Communication, Phys. Rev. Lett. 83,
1874 (1999)

Braunstein S. L., Caves C. M., Jozsa R., Linden N., Popescu S. and Schack
R., Separability of Very Noisy Mixed States and Implications for NMR
Quantum Computing, Phys. Rev. Lett. 83, 1054 (1999)

Braunstein S. L., Mann A. and Revzen M., Maximal Violation of Bell In-
equalities for Mixed States, Phys. Rev. Lett. 68, 3259 (1992)

Brukner C., Pan J.-W., Simon C., Weihs G. and Zeilinger A., Probabilistic
Instantaneous Quantum Computation, Phys. Rev. A 67, 034304 (2003)

Bruß D., Characterizing Entanglement, J. Math. Phys. 43, 4237 (2002)

Buhrman H., Cleve R. and van Dam W., Quantum Entanglement and Com-
munication Complexity, SIAM Journ. on Computing 30, 1829 (2001)

Childs A. M., Leung D., Verstraete F. and Vidal G., Asymptotic entangle-
ment capacity of the Ising and anisotropic Heisenberg interactions, Quan-
tum Information and Computation 3, 97 (2003)

Chuang I. L. and Yamamoto Y., A Simple Quantum Computer, Phys. Rev.

A 52, 3489 (1995)

Cleve R. and Buhrman H., Substituting Quantum Entanglement for Com-

munication, Phys. Rev. A 56, 1201 (1997)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 526

526 Bibliography

Cleve R., Ekert A., Macchiavello C. and Mosca M., Quantum Algorithms
Revisited, Proc. Roy. Soc. Lond. A 454, 339 (1998)

Dehaene J. and De Moor B., Clifford group, stabilizer states, and linear
and quadratic operations over GF(2), Phys. Rev. A 68, 042318 (2003)

de Almeida N. G., All-quantum teleportation, arXiv:1703.07278

Deutsch D., Quantum Computation Networks, Proc. R. Soc. London A

425, 73 (1989)

Deutsch D., Ekert A., Josza R., Macchiavello C., Popescu S. and Sanpera
A., Quantum Privacy Amplification and the Security of Quantum Cryptog-
raphy over Noisy Channels, Phys. Rev. Lett. 77, 2818 (1996)

Deutsch D. and Jozsa R., Rapid solutions of problems by quantum compu-

tation, Proc. R. Soc. London 439, 553 (1992)

Edamatsu K., Complete and Deterministic Bell State Measurement Using

Spin Products, arXiv:1612.08578

Ekert A., Quantum Cryptography Based on Bell’s Theorem, Phys. Rev.

Lett. 67, 661 (1991)

Ekert A. and Knight P. L., Entangled quantum systems and the Schmidt
decomposition, American Journal of Physics, 63, 415 (1995)

Fan H., Matsumoto K. and Imai H., Quantify entanglement by concurrence

hierarchy, J. Phys. A: Math. Gen. 36, 4151 (2003)

Gleason A. M., Measures on the closed subspaces of a Hilbert space, Indiana

University Mathematics Journal, 6, 885 (1957)

Gottesman D., Grand Mathematical Challenge for the Twenty-First Cen-

tury and the Millennium, edited by S. J. Lomonaco Jr., (American Mathe-
matical Society, Providence, Rhode Island), 221 (2002)

Gottesman D., A Class of Quantum Error-Correcting Codes Saturating the

Quantum Hamming Bound, Phys. Rev. A 54, 1862 (1996)

Gottesman D., An Introduction to Quantum Error Correction and Fault-

Tolerant Quantum Computation, arXiv:0904.2557v1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 527

Bibliography 527

Gottesman D., An Introduction to Quantum Error Corrections and Fault-

Tolerant Quantum Computation, in “Quantum Information Science and Its
Contributions to Mathematics”, Proceedings of Symposia in Applied Math-
ematics, 68, 13 (2010)

Grabowski J., Kuś M. and Marmo G., Geometry of quantum systems: den-
sity states and entanglement, J. Phys. A: Math. Gen. 38, 10217 (2005)

Grover L., Quantum Mechanics helps searching for a needle in a haystack,

Phys. Rev. Lett. 79, 325 (1997)

Grudka A. and Wojcik A., How to encode the states of two non-entangled
qubits in one qutrit, Phys. Lett. A 314, 350 (2003)

Hardy Y., Steeb W.-H. and Stoop R., Fully Entangled Quantum States in
2
C N and Bell Measurement, Int. J. Theor. Phys. 42, 2314 (2003)

Hardy Y., Steeb W.-H. and Stoop R., Entanglement, Disentanglement and
Wigner Functions, Physica Scripta 69, 166 (2004)

Hardy Y. and Steeb W.-H., Unitary Operators, Entanglement, and Gram-

Schmidt Orthogonalization, Int. J. of Mod. Phys. C 20, 891 (2009)

Hardy Y. and Steeb W.-H., Genetic Algorithms and Optimization Problems

in Quantum Computing, Int. J. Mod. Phys. C 21, 1359 (2010)

Henderson L. and Vedral V., Information, Relative Entropy of Entangle-

ment and Irreversibility, Physical Review Letters, 84, 2263 (2000)

Horodecki M., P. Horodecki and R. Horodecki, Separability of Mixed States:

necessary and sufficient conditions, Physics Letters A, 223, 1 (1996)

Jordan T. F., Quantum mysteries explored, Am. J. Phys. 62, 874 (1994)

Jozsa R., Quantum Algorithms and the Fourier transform, Proc. Roy. Soc.
London A 454, 323 (1998)

Kielpinski D., A small trapped-ion quantum register, J. Opt. B: Quantum

Semiclass. Opt. 5, R121 (2003)

Kimura G., The Bloch vector for N-level systems, Phys. Lett. A 314, 319
(2003)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 528

528 Bibliography

Korbelář M. and Tolar J., Symmetries of the finite Heisenberg group for
composite systems, J. Phys. A: Math. Theor. 43, 375302 (2010)

Kuzmich A., Walmsley A. and Mandel L., Violation of a Bell-type inequal-

ity in the homodyne measurement of light in an Einstein-Podolsky-Rosen
state, Phys. Rev. A 64, 063804 (2001)

Laflamme L., Miquel C., Paz J. P. and Zurek W. H., Perfect Quantum Er-
ror Correction Code, Phys. Rev. Lett. 77, 198 (1996)

Maassen H. and Uffink J. B. M., Generalized Entropic Uncertainty Rela-

tions, Phys. Rev. Lett. 60, 1103 (1988)

Mermin N. D., From Classical State-Swapping to Quantum Teleportation,

Phys. Rev. A 65, 012320 (2002)

Nielsen M. A., Quantum information theory, Ph.D. thesis, University of

New Mexico, (1998)

Nielsen M. A., Dawson C. M., Dodd J. L., Gilchrist A., Mortimer D.,
Osborne T. J., Bremner M. J., Harrow A. W. and Hines A., Quantum dy-
namics as a physical resource, Physical Review A, 67, 052301 (2003)

Nielsen M. A. and Chuang L., Programmable Quantum Gate Arrays, Phys.

Rev. Lett. 79, 321 (1997)

Pan J.-W., Bouwmeester D., Daniell M., Weinfurter H. and Zeilinger A.,
Experimental test of quantum nonlocality in three-photon Greenberger-Horne-
Zeilinger entanglement, Nature 403, 515 (2000)

Planat M. and Jorrand P., Group theory for quantum gates and quantum
coherence, J. Phys. A: Math. Theor. 41, 182001 (2008)

Sharma S. S., Tripartite GHZ state generation with trapped ion in optical
cavity, Phys. Lett. A 311, 111 (2003)

Shor P. W., Scheme of reducing decoherence in quantum computer memory,

Phys. Rev. A 52, 2493 (1995)

Shor P. W., Polynomial-time algorithm for prime factorization and discrete

logarithms on a quantum computer, Soc. Ind. Appl. Math. J. Comp. 1484
(1997)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 529

Bibliography 529

Steeb W.-H. and Hardy Y., Entangled Quantum States and a C++ Imple-
mentation, Int. J. Mod. Phys. C 11, 69 (2000)

Steeb W.-H. and Hardy Y., Quantum Computing and SymbolicC++ Sim-
ulations, Int. J. Mod. Phys. C 11, 323 (2000)

Steeb W.-H. and Hardy Y., Energy Eigenvalue Level Motion and a Sym-
bolicC++ Implementation, Int. J. Mod. Phys. C 11, 1347 (2000)

Steeb W.-H. and Hardy Y., Fermi Systems, Hubbard Model and a Symbol-
icC++ Implementation, Int. J. Mod. Phys. C 12, 235 (2001)

Steeb W.-H. and Hardy Y., Energy Eigenvalue Level Motion with Two Pa-
rameters, Z. Naturforsch. 56 a, 565 (2001)

Steeb W.-H. and Hardy Y., Entangled Quantum States and the Kronecker
Product, Z. Naturforsch. 57 a, 689 (2002)

Steeb W.-H., Hardy Y., Stoop R., Discrete wavelets and perturbation the-
ory, J. Phys A: Math. Gen. 36, 6807 (2003)

Steeb W.-H. and Hardy Y., A hierarchy of Hamilton operators and entan-
glement, Central European Journal of Physics, 7, 854 (2009)

Steeb W.-H., Diabolic Points and Entanglement, Physica Scripta 81, 025012
(2010)

Steeb W.-H., Hamilton Operators, Discrete Symmetries, Brute Force and

SymbolicC++, Int. J. Mod. Phys. C 24, 1250095 (2013)

Sutter D., Scholz V. B., Winter A. and Renner R., Approximate Degradable
Quantum Channels, arXiv:1412.0980v3

Thomsen M. K., Glück R. and Axelsen H. B., Reversible arithmetic logic

unit for quantum arithmetic, J. Phys. A: Math. Theor. 43, 382002 (2010)

Trifonov A., Tsegaye T., Björk G., Söderholm J., Goobar E., Atatüre M.
and Sergienko A. V., Experimental demonstration of the relative phase op-
erator, J. Opt. B: Quantum Semiclass. Opt., 2, 105 (2000)

Tyson J., Operator-Schmidt decomposition of the quantum Fourier trans-

form on C N1 ⊗ C N2 , J. Phys. A: Math. Gen. 36 6813 (2003)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 530

530 Bibliography

van Loock P. and Braunstein S. L., Greenberger-Horne-Zeilinger nonlocal-

ity in phase space, Phys. Rev. A 63, 022106 (2001)

Wang X. and Sanders B. C., Entanglement capability of self-inverse Hamil-

tonian evolution, Phys. Rev. A 68, 014301 (2003)

Werner R. F., Quantum states with Einstein-Podolsky-Rosen correlations

admitting a hidden-variable model, Phys. Rev. A 40, 4277 (1989)

Wong A. and Christensen N., Potential multiparticle entanglement mea-

sure, Phys. Rev. A 63, 044301

Yang G., Hung W. N. N., Song X. and Perkowski M., Majority based re-
versible logic gates, Theoretical Computer Science 334, 259 (2005)

Yura F., Entanglement cost of three-level antisymmetric states, J. Phys. A:

Math. Gen. 36, L237 (2003)

Part II

Ban M., The phase operator in quantum information processing, J. Phys.

A: Math. Gen. 35, L193 (2002)

Ban M., Phase-space approach to continuous variable quantum teleporta-

tion, Phys. Rev. A 69, 054304 (2004)

Banerji J., Non-linear wave packet dynamics of coherent states, PRAMANA

J. Phys. 56, 267 (2001)

Barak R. and Ben-Aryeh Y., Photon statistics and entanglement in coherent-

squeezed linear Mach-Zehnder and Michelson interferometers, J. Opt. Soc.
Am. B 25, 361 (2008)

Bartlett S. D., Sanders B. C., Braunstein S. L. and Nemoto K., Efficient

Classical Simulations of Continuous Variable Quantum Information Pro-
cesses, Phys. Rev. Lett. 88, 097904 (2002)

Barnett S. M., Jeffers J. and Gatti A., Quantum optics of lossy beam split-
ters, Phys. Rev. A 57, 2134 (1998)

Braunstein S. L. and Kimble H. J., Teleportation of Continuous Quantum

Variables, Phys. Rev. Lett. 80, 869 (1998)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 531

Bibliography 531

Braunstein S. L., Fuchs C. A., Kimble H. J. and van Loock P., Quantum
versus classical domains for teleportation with continuous variables, Phys.
Rev. A 64, 022321 (2001)

Buzek V., Vidiella-Barranco and Knight P. L., Superpositions of coherent

states: Squeezing and dissipation, Phys. Rev. A 45, 6570 (1992)
Carmichael H. J., Milburn G. J. and Walls D. F., Squeezing in a detuned
parametric amplifier, J. Phys. A: Math. Gen. 17, 469 (1984)

Cerf N. J., Adami C. and Kwiat P. G., Optical Simulation of Quantum

Logic, Phys. Rev. A 57, 1477 (1998)

Cirac J. I., Quantum Information: Entanglement, Purification, Error Cor-

rection, and Quantum Optical Implementations, in Fundamentals of Quan-
tum Information, Heiss D. (Ed.), Springer, Berlin (2002)

Chruściński D., Spectral Properties of the Squeeze Operator, Phys. Lett. A

327, 290-295 (2004)

D’Ariano G. M., De Martin F. and Sacchi M. F., Continuous Variable

Cloning via Network of Parametric Gates, Phys. Rev. Lett. 86, 914 (2001)

DasGupta A., Disentanglement formulas: An alternative derivation and

some applications to squeezed coherent states, Am. J. Phys. 64, 1422 (1996)

de Oliveira M. C. and Milburn G. J., Discrete teleportation protocol of con-

tinuum spectra field states, Phys. Rev. A 65, 032304 (2002)

Dell’Anno F., De Siena S. and Illuminati F., Multiphoton quantum optics

and quantum state engineering, Physics Reports, 428, 53 (2006)

Dodonov V. V., Man’ko V. I. and Nikonov D. E., Even and odd coherent
states for multimode parametric systems, Phys. Rev. A 51, 3328 (1995)

Einstein A., Podolsky B. and Rosen N., Can Quantum-Mechanical De-

scription of Physical Reality Be Considered Complete?, Phys. Rev. 47,
777 (1935)

Eisert J., Optimizing Linear Optics Quantum Gates, Phys. Rev. Lett. 95,
040502 (2005)

Ferraro A., Olivares S. and Paris M. G. A., Gaussian states in a continuous

variable quantum information, arXiv:quant-ph/0503237v1, 2005
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 532

532 Bibliography

Fujii K., Basic Properties of Coherent and Generalized Coherent Operators

Revisited, Mod. Phys. Lett. A 16, 1277 (2001)

Hardy Y. and Steeb W.-H., Fermi-Bose Systems, Macroscopic Quantum

Superposition States and Entanglement, Int. J. Theor. Phys. 43, 2207
(2004)

Hong-Yi F. and Klauder J. R., Canonical coherent state representation of

some squeeze operators, J. Phys. A: Math. Gen. 21, L725-L730 (1988)

Imoto N., Haus H. A. and Yamamoto Y., Quantum nondemolition mea-

surement of the photon number via the optical Kerr effect, Phys. Rev. A
32, 2287 (1985)

Jeffers J., Barnett S. M. and Pegg D., Retrodiction as a tool for micromaser
field measurement, J. Mod. Opt. 49, 925 (2002)

Joo J. and Ginossar E., Hybrid teleportation via entangled coherent states
in circuit quantum electrodynamics, arXiv:1509.02859

Knill E., Laflamme R. and Milburn G. J., A scheme for efficient quantum
computation with linear optics, Nature 409, 46 (2001)

Kok P., Munro W. J., Nemoto K., Ralph T. C., Dowling J. P. and Milburn
G. J., Linear optical quantum computing with photonic qubits, Rev. Mod.
Phys. 79, 1 (2007)

Koniorczyk M., Buzek V. and Janszky J., Wigner-function description

of quantum teleportation in arbitrary dimensions and a continuous limit,
Phys. Rev. A 64, 034301 (2001)

Kok P., Munro W. J., Nemoto K., Ralph T. C., Dowling J. P. and Milburn
G. J., Linear Optical Quantum Computing, arXiv:quant-ph/0512071v2

Kok P. and Braunstein S. L., Multi-dimensional Hermite polynomials in

quantum optics, J. Phys. A: Math. Gen. 34, 6185 (2001)

Lee J. and Kim M. S., Quantum Teleportation Via Mixed Two-Mode Squeezed
States in the Coherent-State Representation, J. Kor. Phys. Soc. 42, 457
(2003)

Li Shang-Bin and Xu Jing-Bo, Quantum probabilistic teleportation via en-

tangled coherent states, Phys. Lett. A 309, 321 (2003)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 533

Bibliography 533

Lloyd S. and Braunstein S. L., Quantum Computing over Continuous Vari-

ables, Phys. Rev. Lett. 82, 1784 (1999)

Lo C. F., Eigenfunctions and eigenvalues of squeeze operators, Phys. Rev.

A 42, 752 (1990)

Milburn G. J. and Braunstein S. L., Quantum teleportation with squeezed

states, Phys. Rev. A 60, 937 (1999)

Pan Jian-Wei, Daniell M., Gasparoni S., Weihs G., and Zeilinger A., Ex-
perimental Demonstration of Four-Photon Entanglement and High-Fidelity
Teleportation, Phys. Rev. Lett. 86, 4435 (2001)

Paris M. G. A., Generation of mesoscopic quantum superpositions through

Kerr-stimulated degenerate downconversion, J. Opt. B: Quantum Semi-
class. Opt. 1, 662 (1999)

Pegg D. T., Barnett S. M. and Jeffers J., Quantum theory of preparation

and measurement, J. Mod. Opt. 49, 913 (2002)

Ralph T. C., Munro W. J. and Milburn G. J., Quantum Computation with

Coherent States, Linear Interactions and Superposed Resources, arXiv:quant-
ph/0110115v2

Schrödinger E., Die gegenwärtige Situation in der Quantenmechanik, Natur-

wissenschaften 23, pp. 807-812; 823-828; 844-849 (1935)

Serafini A., De Siena S., Illuminati F. and Paris M. G. A., Minimum deco-
herence cat-like states in Gaussian noisy channels, arXiv:quant-ph/0310005v2

Sharma S. S. and Vidiella-Barranco A., Fast quantum logic gates with

trapped ions interacting with external laser and quantized cavity field be-
yond Lamb-Dicke regime, Phys. Lett. A 309 345 (2003)

Slosser J. J. and Milburn G. J., Creating Metastable Schrödinger Cat States,

Phys. Rev. Lett. 75, 418 (1995)

Steeb W.-H., Entanglement, Bose operators, coherent states and classical

dynamical systems, J. Phys. A: Math. Theor. 40, 9577 (2007)

Steeb W.-H. and Hardy Y., Quantum Optics Networks, Unitary Operators
and Computer Algebra, Int. J. of Mod. Phys. C 19, 1069 (2008)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 534

534 Bibliography

Steeb W.-H., Ordinary Differential Equations, First Integrals, Conformal

Invariants, Bose Operators, Coherent States and Entanglement, Int. J.
Geom. Methods in Modern Physics, 5, 1057 (2008)

Thamm R., Kolley E. and Kolley W., Green’s Functions for Tight-Binding
Electrons on a Square Lattice in a Magnetic Field, phys. stat. sol. (b) 170,
163 (1992)

van Enk S. J. and Hirota O., Entangled coherent states: Teleportation and
decoherence, Phys. Rev. A 64, 022313 (2001)

van Loock P. and Braunstein S. L., Multipartite Entanglement for Contin-

uous Variables: A Quantum Teleportation Network, 84, 3482 (2000)

Vaziri A., Mair A., Weihs G. and Zeilinger A., Entanglement of the Angular
Orbital Momentum States of the Photons, Nature 412, 313 (2001)

Youn S.-H., Conditional Generation Scheme for Entangled Vacuum Evac-

uated Coherent States by Mixing Two Coherent Beams with a Squeezed
Vaccum State, arXiv:1511.07956

Yurke B. and Stoler D., Generating quantum mechanical superpositions of

macroscopically distinguishable states via amplitude dispersion, Phys. Rev.
Lett. 57, 13 (1986)

Zou Xu-Bo, Pahlke K. and Mathis W., The non-deterministic quantum logic
operation and the teleportation of the superposition of the vacuum state and
the single-photon state via parametric amplifiers, Phys. Lett. A 311 271
(2003)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 535

Index
C ∗ -algebra, 195 Bilateral XOR, 366
SL(2, C), 46 Bit-flip error, 352
SO(4), 177 Bloch coherent states, 112
SU (1, 1), 453 Bloch representation, 518
SU (2), 85, 452 Bloch sphere, 237
SU (N ), 86 Bloch vector, 133
π-meson, 59 Bogolubov transform, 396, 401
su(1, 1), 402, 418, 447 Bogolubov unitary operator, 402
su(2), 403 Bose annihilation operator, 387
su(N ), 86 Bose creation operator, 387
Bose-Einstein condensates, 511
Abelian group, 188 Bose-Einstein density operator, 396,
Algorithm, 325 455
AND gate, 155 Braid like relation, 202
AND operation, 327 Bures distance, 18, 120
Anti-commutation relations, 276 Bures metric, 274
Anticommutator, 9
Antilinear operator, 95 Canonical, 402
Antisymmetric subspace, 263 Cauchy integral formula, 108
Associated Laguerre polynomials, 415 Cayley-Hamilton theorem, 16, 66, 74,
88
Baker-Campbell-Hausdorff formula,
Characteristic length, 389
76, 84, 428
Charge conjugation, 94
Beam splitter, 409, 414, 478, 484,
Chebyshev polynomials, 411
501
Choi-Jamiolkowski representation, 378
Beam splitter interaction, 433
CHSH inequality, 303
Beam splitter operator, 478, 479
Circulant matrix, 78
Bell angles, 303
Classical algorithm, 325
Bell basis, 35, 201, 262, 309, 312
Classical information capacity, 215
Bell inequality, 296, 297
Clone, 497
Bell matrix, 59
Cloning, 319
Bell operator, 299, 301, 306
CNOT gate, 55, 197, 320
Bell state, 43, 253, 256
Coherent squeezed state, 454
Bell states, 39, 44, 130, 197, 199,
Coherent states, 272, 423
252, 282, 473
Comb, 95
Bell-CHSH inequality, 484
Communication complexity, 329, 331
Bijection, 24
Commutative group, 188
Bilateral exclusive or, 366

535
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 536

536 Index

Comparison method, 90 Expectation value, 115

Completely positive map, 372
Completeness relation, 6, 112, 183, Factorisability, 297
185, 391, 427, 462 Fermi operators, 276
Concurrence, 265, 266 Feynman gate, 157
Continuity argument, 69 Fidelity, 274
Controlled NOT, 179 Fock states, 388
Convex combination, 131 Fourier matrix, 79, 176
Correlation, 291 Fourier transform, 420, 499
Cosine-sine decomposition, 66 Fredkin gate, 160, 162, 193, 409
Critical points, 217
Cross phase modulation, 507 Gamma matrices, xiii
Csanky’s algorithm, 74 Gaussian, 429
Cyclic invariance, 38, 87, 133, 141 Gaussian distribution, 420
Gell-Mann matrices, 128
Density matrix, 4, 115, 121, 129, 144 Generalized Fredkin gate, 162
Density matrix purification, 129 Generalized measurement, 244
Density operator, 395 Generalized Toffoli gate, 161
Determinant, 158 Geometric series, 396
Deutsch’s problem, 335, 337 GHZ state, 41, 42, 49, 195, 261, 279,
Dicke states, 48 283, 286
Dirac delta function, 471 Gleason’s theorem, 239
Dirac matrices, 37 Global phase, 238
Direct sum, 106 Gram’s matrix, 111
Disentangle, 490 Graviton, 92
Disentangled, 251 Group, 91, 428
Disentangled form, 90 GXOR gate, 187
Displaced parity operator, 484
Displacement operator, 422, 423 Haar measure, 289
Hadamard basis, 3, 6, 7, 143
Eigenvalue equation, 65 Hadamard gate, 89, 196
Eigenvalue kick back, 332 Hadamard matrix, 79
Electric field, 414, 444 Hadamard operator, 8
Entangled, 251, 253 Hamming weight, 338
Entanglement, 474 Harmonic oscillator, 389
Entanglement capability, 267, 270 Heisenberg algebra, 387, 423, 435
Entanglement of formation, 265 Heisenberg commutation relation, 76
Entanglement witness, 283 Heisenberg equation of motion, 11,
Entire analytic function, 397, 427 43, 509
Entropic uncertainty relation, 210, Heisenberg-Weyl group, 408
222 Heterodyne-current operator, 503
EPR state, 148, 298, 471 Hilbert space, xiii
Error syndrome, 352 Hilbert-Schmidt distance, 120
Euler’s identity, 184 Hilbert-Schmidt norm, 37, 65
Expectation, 236
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 537

Index 537

Homodyne detector, 502 NAND gate, 155

Hubbard model, 276 No cloning theorem, 319, 321
Husimi distribution, 429, 454 Nonlinear coherent states, 431
Hyperdeterminant, 286 NOR gate, 155
Hyperplane, 109 Norm, 37, 67
Normal, 66
Imperfect cloning, 497 Normal matrix, 78, 81, 118
Intensity, 512 Normal order form, 398
Invariant, 46, 286 NOT gate, 13, 81, 155
Inversion about average, 185 NOT operation, 6, 179–181
Number operator, 388, 391, 405
Kerr medium, 430, 508 Number state, 508
Kerr-Hamilton operator, 506 Number states, 388
Klein’s inequality, 210, 216
Kramer’s vector, 105 Operator-Schmidt decomposition, 54
Kraus map, 371 OR gate, 155
Kraus operators, 371, 373 Orthogonal measurement, 241
Kraus representation, 378
Kronecker product, 32 Parameter differentiation, 83, 392, 400,
Kullback-Leibler distance, 221 410
Parity function, 327
Lagrange multiplier method, 67 Parity operator, 411
Lagrange multiplier problem, 217 Parseval’s relation, 52
Lamb-Dicke parameters, 516 Partial measurement, 231
Landau gauge, 518 Partial trace, 142, 477
Levi-Civita symbol, 264 Partial transpose, 373
Lie algebra, 408 Partition function, 50
Log-sum inequality, 221 Pauli group, 408
Logarithm, 197 Pauli spin matrices, xii, 4, 182
Lorentz metric, 25 Permutation matrices, 176
Lorentzian distribution, 420 Permutation matrix, 76
Phase change transform, 237
Mach-Zehnder interferometer, 12 Phase flip error, 352
Macroscopic quantum superposition, Phase modulator, 409
481 Phase operator, 199
Magic gate, 177 Phase shift, 405
Maximally entangled, 252, 281 Photon, 105
Maxwell’s equations, 105 Poisson distribution, 418, 420, 426,
Mixed state, 4, 115 430
Modified Bures metric, 274 Polar decomposition theorem, 69
Momentum boost operator, 408 Polarization vector, 414
Mutually unbiased, 93, 101 Polarizing beam splitter, 282
Mutually unbiased bases, 23 Position eigenstates, 408
Mutually unbiased basis, 100 Positive map, 372
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 538

538 Index

Positive operator-valued measure, 228 Shannon information entropy, 210,

Positive semidefinite, 81, 115 222
POVM, 228 Similar matrices, 69
Primary permutation matrix, 78 Singlet state, 307
Probability, 236 Singular value decomposition, 50
Product state, 251 Skew-hermitian matrix, 70
Projection matrices, 18 Spectral decomposition, 66
Projective measurement, 236 Spectral representation, 135, 236
Pure state, 4, 115, 129, 209 Spherical coordinates, 323
Putzer method, 88 Spin coherent state, 98
Spin flipped density matrix, 123
Quantum 2-torus, 195 Spontaneous parametric down con-
Quantum algorithm, 325 version, 474
Quantum correlation function, 68, 249 Square root of NOT, 181
Quantum Fourier transform, 183, 326 Squeezing operator, 440, 442, 444
Quantum relative entropy, 210, 214, Squeezing parameter, 396, 440
216 Stabilized, 47
Qubit, 3 Stabilizer, 355
Qubit trine, 9 Standard basis, 3, 6, 7
Qudits, 187 State entanglement rate, 267, 270
Qutrit state, 240 Statistical independence, 297
Stinespring representation, 380
Rabi frequency, 510
Subadditivity, 218
Reduced density matrix, 289
Supercomplete, 427
Relative entropy of entanglement, 213
Superposition, 3
Remoteness, 288
Susskind-Glogower canonical phase
Repeated commutator, 446
states, 399
Resolution of identity, 273
Swap gate, 201, 203
Riccati equation, 401
Swap operator, 204, 495
Scalar product, 37, 77 Symmetric group, 181
Schmidt angle, 288
Schmidt basis, 464 Tangle, 124, 289
Schmidt decomposition, 50, 130, 200, Technique of parameter differentia-
260, 268 tion, 73
Schmidt number, 259 Teleported, 309
Schmidt rank, 53, 260 Tensor product, 31
Schrödinger cat states, 432, 434 Tetrahedron, 125
Schrödinger equation, 11, 131 Three tangle, 286
Schur’s theorem, 126 Toffoli gate, 160, 174, 192
Schwinger two-bosons realization, 476 Tomography, 125
Second harmonic generation, 509 Trace, 141, 394, 463
Second law of thermodynamics, 322 Trace distance, 120
Separable, 251 Transfer operator, 414
Shannon entropy, 266 Transmissivity, 414
January 30, 2018 15:21 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 539

Index 539

Transpose, 372 Von Neumann equation, 132, 267

Triangle inequality, 127 Von Neumann measurement, 235
Trotter formula, 113
Truth table, 155 W state, 49, 241, 280, 287
Twin beam state, 433, 453 Walsh-Hadamard gate, 183
Two mode Bell function, 484 Walsh-Hadamard transform, 7, 34,
Two mode squeezed vacuum, 453 178, 337
Werner state, 122, 265, 266
Uhlmann’s transition probability, 274 Weyl expansion, 456
Unary gates, 25 Weyl representation, 76
Uncertainty relation, 136, 450 Wigner function, 147, 148, 471
Undisentangled form, 90 Wigner operator, 147
Unitary depolarizers, 279
Unitary operator, 175 XOR gate, 155, 186
Universal gates, 155 XOR operation, 327
XY-model, 189
Vacuum state, 387
Variance, 19, 26, 116, 288, 425, 472 Yang-Baxter equation, 107
Vec operator, 369 Yurke-Stoler states, 434
Vector product, 17
Von Neumann entropy, 209, 256, 265, Zassenhaus formula, 90
267, 270, 322