Discrete-Time Signals and Systems

Introduction to Digital Signal Processing

Reference:
Sections 1.1 - 1.4 of

Professor Deepa Kundur

John G. Proakis and Dimitris G. Manolakis, Digital Signal Processing:

Principles, Algorithms, and Applications, 4th edition, 2007.

University of Toronto

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

1 / 51

1.1 Signals, Systems and Signal Processing

Chapter 1: Introduction

What is a Signal? What is a System?

I

2 / 51

1.1 Signals, Systems and Signal Processing

Independent Variable

Signal:
I

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

any physical quantity that varies with time, space, or any other
independent variable or variables
Examples: pressure as a function of altitude, sound as a
function of time, color as afunction of space, . . .
x(t) = cos(2t), x(t) = 4 t + t 3 , x(m, n) = (m + n)2

A signal can be represented as a function x(t) and consists of:

1. one or more dependent variable components (e.g., air pressure
x, R-G-B color [x1 x2 x3 ]T );
2. one or more independent variables (e.g., time t, 3-D spacial
location (s1 , s2 , s3 )).

System:
I
I

a physical device that performs an operation on a signal

Examples: analog amplifier, noise canceler, communication
channel, transistor, . . .
y (t) = 4x(t), dydt(t) + 3y (t) = dx(t)
dt + 6x(t),
y (n) 12 y (n 2) = 3x(n) + x(n 2)

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Please note: in this course we will typically use time t to

represent the independent variable although in general it can
correspond to any other type of independent variable.

3 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

4 / 51

Chapter 1: Introduction

1.2 Classification of Signals

Chapter 1: Introduction

Continuous-Time versus Discrete-Time Signals

I

Continuous-Time versus Discrete-Time Signals

I

Continuous-Time Signals: signal is defined for every value of

time in a given interval (a, b) where a and b .
I

1.2 Classification of Signals

Examples: voltage as a function of time, height as a function of

pressure, number of positron emissions as a function of time.

Discrete-Time Signals: signal is defined only for certain specific

values of time; typically taken to be equally spaced points in an
x(t)
x(t)
interval.
2

I
-4

Examples: number of stocks traded per day, average income per

t
t
province.
1
0.5

-3

-2

-1

-4

-3

-2

-1

0.5

-2

x(t)

-3

-2

1
0.5
-4

-3

-2

-1

0.5

-2

-1

1.5

2.5

t
-3

Chapter 1: Introduction

1
-3

-2

-1

-2

-1

1
-3

-2

-1

n
5 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

1.2 Classification of Signals

Chapter 1: Introduction

Continuous-Amplitude versus Discrete-Amplitude

6 / 51

1.2 Classification of Signals

Continuous-Amplitude versus Discrete-Amplitude

x[n]

x(t)

Continuous-Amplitude Signals: signal amplitude takes on a

spectrum of values within one or more intervals
I

x[n]

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

-2

2.5

-2

x[n]

-3

x[n]

x[n]

-1

1.5

x(t)

-4

-2

1
Discrete-Amplitude Signals: signal amplitude takes
on values
from a finite set
2

-4

Examples: color, temperature, pain-level

-3

-2

-1

-3

-2

-1

Examples: digital image, population of a country

-2

x[n]

x(t)

-4

-3

-2

x(t)

-1

-3

-2

-1

x(t)

-1

0.5

-2

-2

-1

0.5

1.5

2.5

1
-3

-2

-1

1
3

x[n]

1
0.5
-2

-3

-3

1
0.5
-4

-2

-2

-4

x[n]
2

1.5

2.5

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

1
-3

-2

-1

1
3

n
7 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

8 / 51

Chapter 1: Introduction

1.2 Classification of Signals

Chapter 1: Introduction

Analog and Digital Signals

I
I

Analog and Digital Signals

analog signal = continuous-time + continuous amplitude

digital signal = discrete-time + discrete amplitude
continuous amplitude

-3

-2

1
0.5
-4

-3

-2

-1

0.5

-2

-1

1.5

2.5

Digital signals are important because they facilitate the use of

digital signal processing (DSP) systems, which have practical
and performance advantages for several applications.

x[n]
2

-2

-2

x[n]

-3

-1

discrete-time

Analog signals are fundamentally significant because we must

interface with the real world which is analog by nature.

x(t)

-4

I
discrete amplitude

x(t)

continuous-time

1
-3

-2

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

-1

1
3

9 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

1.2 Classification of Signals

Chapter 1: Introduction

Analog and Digital Systems

analog system =

advantages: easy to interface to real world, do not need A/D or

D/A converters, speed not dependent on clock rate

digital system =
digital signal input + digital signal output
I

advantages: re-configurability using software, greater control

over accuracy/resolution, predictable and reproducible behavior

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

1.2 Classification of Signals

Deterministic signal:

analog signal input + analog signal output

I

10 / 51

Deterministic vs. Random Signals

I

1.2 Classification of Signals

11 / 51

any signal that can be uniquely described by an explicit

mathematical expression, a table of data, or a well-defined rule
past, present and future values of the signal are known precisely
without any uncertainty

Random signal:
I

I
I

any signal that lacks a unique and explicit mathematical

expression and thus evolves in time in an unpredictable manner
it may not be possible to accurately describe the signal
the deterministic model of the signal may be too complicated to
be of use.

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

12 / 51

Chapter 1: Introduction

1.3 The Concept of Frequency

Chapter 1: Introduction

What is a pure frequency signal?

1.3 The Concept of Frequency

What is a pure frequency signal?

xa (t) = A cos(t + ) = A cos(2Ft + ), t R

I
I
I

I
I

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

13 / 51

1.3 The Concept of Frequency

analog signal, A xa (t) A and < t <

A = amplitude
= frequency in rad/s
F = frequency in Hz (or cycles/s); note: = 2F
= phase in rad

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

Continuous-time Sinusoids

14 / 51

1.3 The Concept of Frequency

Continuous-time Sinusoids: Frequency

xa (t) = A cos(t + ) = A cos(2Ft + ), t R

1. for F R, xa (t) is periodic
I

i.e., there exists Tp

R+

such that xa (t) = xa (t + Tp )

smaller F , larger T

2. distinct frequencies result in distinct sinusoids

I

i.e., for F1 6= F2 , A cos(2F1 t + ) 6= A cos(2F2 t + )

3. increasing frequency results in an increase in the rate of

oscillation of the sinusoid
I

i.e., for |F1 | < |F2 |, A cos(2F1 t + ) has a lower rate of

oscillation than A cos(2F2 t + )

larger F , smaller T
Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

15 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

16 / 51

Chapter 1: Introduction

1.3 The Concept of Frequency

Chapter 1: Introduction

Discrete-time Sinusoids

Discrete-time Sinusoids

x(n) = A cos(n + ) = A cos(2fn + ), n Z

x(n) = A cos(n + ) = A cos(2fn + ), n Z

I
I
I
I
I

discrete-time signal (not digital), A xa (t) A and n Z

A = amplitude
= frequency in rad/sample
f = frequency in cycles/sample; note: = 2f
= phase in rad

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

1.3 The Concept of Frequency

17 / 51

1. x(n) is periodic only if its frequency f is a rational number

I
I

Note: rational number is of the form kk21 for k1 , k2 Z

periodic discrete-time sinusoids:

x(n) = 2 cos( 47 n), x(n) = sin( 5 n + 3)

aperiodic discrete-time sinusoids:

x(n) = 2 cos( 47 n), x(n) = sin( 2n + 3)

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

1.3 The Concept of Frequency

Chapter 1: Introduction

18 / 51

1.3 The Concept of Frequency

Discrete-time Sinusoids
MINIMUM OSCILLATION

x(n) = A cos(n + ) = A cos(2fn + ), n Z

2. radian frequencies separated by an integer multiple of 2 are
identical
I

or cyclic frequencies separated by an integer multiple are

identical
MAXIMUM OSCILLATION

3. lowest rate of oscillation is achieved for = 2k and highest

rate of oscillation is achieved for = (2k + 1), for k Z
I

subsequently, this corresponds to lowest rate for f = k (integer)

and highest rate for f = 2k+1
(half integer), for k Z.
2

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

19 / 51

MINIMUM OSCILLATION

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

20 / 51

Chapter 1: Introduction

1.3 The Concept of Frequency

Chapter 1: Introduction

Complex Exponentials

Complex Exponentials

e j = cos() + j sin()
cos() =

e j +e j
2

sin() =

e j e j
2j

Eulers relation
Continuous-time: A e j(t+) = A e j(2Ft+)
Discrete-time:

where j ,

21 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

1.3 The Concept of Frequency

Chapter 1: Introduction

Periodicity: Continuous-time

22 / 51

1.3 The Concept of Frequency

Periodicity: Discrete-time

x(t + T ), T R+
A e j(2F (t+T )+)
e j2Ft e j2FT e j
e j2FT
e j2FT , k Z
k
T =
k Z
F
1
T0 =
, k = sgn(F )
|F |

x(n + N), N Z+
A e j(2f (n+N)+)
e j2fn e j2fN e j
e j2fN
e j2fN , k Z
k
=
k Z
N
k0
k0
=
, min |k 0 | Z such that
Z+
f
f

x(n) =
j(2fn+)
Ae
=
j2fn
j
e
e
=
1 =
j2k
e
=1 =

x(t) =
j(2Ft+)
Ae
=
j2Ft
j
e
e
=
1 =
j2k
e
=1 =

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

A e j(n+) = A e j(2fn+)

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

1.3 The Concept of Frequency

f
N0

23 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

24 / 51

Chapter 1: Introduction

1.3 The Concept of Frequency

1.3 The Concept of Frequency

1
6

Example 1: = /6 =

x[n] = cos

Chapter 1: Introduction

 n 
6

2k
2k
= 1 = 12k

6
= 12 for k = 1

N =
N0

The fundamental period is 12 which corresponds to k = 1 envelope

cycles.

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

1.3 The Concept of Frequency

Example 2: = 8/31 =

x[n] = cos

25 / 51

ENVELOPE CYCLES

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

26 / 51

1.3 The Concept of Frequency

8
31

8n
31

2k
2k
31
= 8 = k

4
31
= 31 for k = 4

N =
N0

The fundamental period is 31 which corresponds to k = 4 envelope

cycles.

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

27 / 51

ENVELOPE CYCLES

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

28 / 51

Chapter 1: Introduction

1.3 The Concept of Frequency

Example 3: = 1/6 =

x[n] = cos

Chapter 1: Introduction

1.3 The Concept of Frequency

1
6

n
6
NOT PERIODIC

N Z+

2k
2k
= 1 = 12k

6
does not exist for any k Z; x[n] is non-periodic.

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

29 / 51

1.3 The Concept of Frequency

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

Uniqueness: Continuous-time

30 / 51

1.3 The Concept of Frequency

Uniqueness: Discrete-time
Let f1 = f0 + k where k Z,

For F1 6= F2 ,
A cos(2F1 t + ) 6= A cos(2F2 t + )
except at discrete points in time.

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

x1 (n) =
=
=
=

31 / 51

A e j(2f1 n+)
A e j(2(f0 +k)n+)
A e j(2f0 n+) e j(2kn)
x0 (n) 1 = x0 (n)

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

32 / 51

Chapter 1: Introduction

-3

-3

-2

-2

1.3 The Concept of Frequency

-1

-1

Chapter 1: Introduction

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

-3

-3

33 / 51

-1

34 / 51

1.3 The Concept of Frequency

Harmonically Related Complex Exponentials

Harmonically related sk (t) = e jk0 t = e j2kF0 t ,
(cts-time)
k = 0, 1, 2, . . .

Therefore, dst-time sinusoids are unique for f [0, 1).

For any sinusoid with f1 6 [0, 1), f0 [0, 1) such that

Example: A dst-time sinusoid with frequency f1 = 4.56 is the

same as a dst-time sinusoid with frequency f0 = 4.56 4 = 0.56.
Example: A dst-time sinusoid with frequency f1 = 78 is the
same as a dst-time sinusoid with frequency f0 = 78 + 1 = 18 .

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

x1 (n) = A e j(2f1 n+) = A e j(2f0 n+) = x0 (n).

-2

-1

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

1.3 The Concept of Frequency

Uniqueness: Discrete-time
I

-2

1.3 The Concept of Frequency

35 / 51

Scientific Designation
C-1
C0
C1
C2
C3
C4
..
.

Frequency (Hz)
8.176
16.352
32.703
65.406
130.813
261.626
..
.

k for F0 = 8.176
1
2
4
8
16
32

C9

8372.018

1024

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

36 / 51

Chapter 1: Introduction

1.3 The Concept of Frequency

Chapter 1: Introduction

Harmonically Related Complex Exponentials

Scientific Designation
C1
C2
C3
C4 (middle C)
C5
C6
C7
C8

C1

C2

C3

Frequency (Hz)
32.703
65.406
130.813
261.626
523.251
1046.502
2093.005
4186.009

C4

C6

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

Harmonically Related Complex Exponentials

k for F0 = 8.176
4
8
16
32
64
128
256
512

C5

1.3 The Concept of Frequency

What does the family of harmonically related sinusoids sk (t) have in

common?
Harmonically related sk (t) = e jk0 t = e j2(kF0 )t ,
(cts-time)
k = 0, 1, 2, . . .

1
1
=
cyclic frequency
kF0
period: Tk = any integer multiple of T0
1
common period: T = k T0,k =
F0
fund. period: T0,k =

C7

C8

37 / 51

1.3 The Concept of Frequency

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

Harmonically Related Complex Exponentials

38 / 51

1.3 The Concept of Frequency

Harmonically Related Complex Exponentials

Discrete-time Case:
For periodicity, select f0 =

1
N

where N Z:

sk+N (n) =
=
=
=

Harmonically related sk (n) = e j2kf0 n = e j2kn/N ,

(dts-time)
k = 0, 1, 2, . . .

Therefore, there are only N distinct dst-time harmonics:

sk (n), k = 0, 1, 2, . . . , N 1.

There are only N distinct dst-time harmonics:

sk (n), k = 0, 1, 2, . . . , N 1.

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

e j2(k+N)n/N
e j2kn/N e j2Nn/N
e j2kn/N 1
e j2kn/N = sk (n)

39 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

40 / 51

Chapter 1: Introduction

1.4 Analog-to-Digital and Digital-to-Analog Conversion

Chapter 1: Introduction

Analog-to-Digital Conversion

Analog-to-Digital Conversion

A/D converter
xa(t)

A/D converter
xq(n)

x(n)
Sampler

Analog
signal

1.4 Analog-to-Digital and Digital-to-Analog Conversion

Quantizer
Discrete-time
signal

xa(t)

01011...
Coder

Quantized
signal

Digital
signal

xq(n)

x(n)
Sampler

Analog
signal

Quantizer
Discrete-time
signal

01011...
Coder

Quantized
signal

Digital
signal

Sampling:
I conversion from cts-time to dst-time by taking samples at
discrete time instants
I E.g., uniform sampling: x(n) = xa (nT ) where T is the sampling
period and n Z

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

41 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

1.4 Analog-to-Digital and Digital-to-Analog Conversion

Chapter 1: Introduction

Analog-to-Digital Conversion

Analog
signal

A/D converter
x(n)

Sampler

1.4 Analog-to-Digital and Digital-to-Analog Conversion

Analog-to-Digital Conversion

A/D converter
xa(t)

42 / 51

Quantizer
Discrete-time
signal

xq(n)

Quantized
signal

xa(t)

01011...
Coder
Digital
signal

Analog
signal

Quantization:
I conversion from dst-time cts-valued signal to a dst-time
dst-valued signal
I quantization error: eq (n) = xq (n) x(n) for all n Z

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

x(n)
Sampler

Quantizer
Discrete-time
signal

xq(n)

Quantized
signal

01011...
Coder
Digital
signal

Coding:
I representation of each dst-value xq (n) by a
b-bit binary sequence
I e.g., if for any n, xq (n) {0, 1, . . . , 6, 7}, then the coder may
use the following mapping to code the quantized amplitude:

43 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

44 / 51

Chapter 1: Introduction

1.4 Analog-to-Digital and Digital-to-Analog Conversion

Chapter 1: Introduction

Analog-to-Digital Conversion

Sampling Theorem

A/D converter
xa(t)

x(n)
Sampler

Analog
signal

Quantizer
Discrete-time
signal

1.4 Analog-to-Digital and Digital-to-Analog Conversion

xq(n)

If the highest frequency contained in an analog signal xa (t) is

Fmax = B and the signal is sampled at a rate

01011...
Coder

Quantized
signal

Fs > 2Fmax = 2B

Digital
signal

then xa (t) can be exactly recovered from its sample values using the
interpolation function

Example coder:
0
1
2
3

000
001
010
011

4
5
6
7

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

g (t) =

100
101
110
111

Note: FN = 2B = 2Fmax is called the Nyquist rate.

45 / 51

1.4 Analog-to-Digital and Digital-to-Analog Conversion

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

Sampling Theorem

46 / 51

1.4 Analog-to-Digital and Digital-to-Analog Conversion

Bandlimited Interpolation

Sampling Period = T =

1
1
=
Fs
Sampling Frequency

Therefore, given the interpolation relation, xa (t) can be written as

xa (t) =

sin(2Bt)
2Bt

x(n) samples

bandlimited interpolation
function -- sinc

xa (nT )g (t nT )

n=
0

xa (t) =

x(n) g (t nT )

n=

where xa (nT ) = x(n); called bandlimited interpolation.

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

47 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

48 / 51

Chapter 1: Introduction

1.4 Analog-to-Digital and Digital-to-Analog Conversion

Chapter 1: Introduction

Digital-to-Analog Conversion

Digital-to-Analog Conversion

x(n)

original/bandlimited
interpolated signal

original/bandlimited
interpolated signal

1.4 Analog-to-Digital and Digital-to-Analog Conversion

Common interpolation approaches: bandlimited interpolation,

zero-order hold, linear interpolation, higher-order interpolation
techniques, e.g., using splines
In practice, cheap interpolation along with a smoothing filter
is employed.

zero-order
hold
I

-3T -2T -T

2T

Common interpolation approaches: bandlimited interpolation,

zero-order hold, linear interpolation, higher-order interpolation
techniques, e.g., using splines
In practice, cheap interpolation along with a smoothing filter
is employed.


Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

Chapter 1: Introduction

49 / 51

Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

1.4 Analog-to-Digital and Digital-to-Analog Conversion

Digital-to-Analog Conversion
original/bandlimited
interpolated signal

-3T -2T -T

linear
interpolation

2T

3T

Common interpolation approaches: bandlimited interpolation,

zero-order hold, linear interpolation, higher-order interpolation
techniques, e.g., using splines
In practice, cheap interpolation along with a smoothing filter
is employed.


Professor Deepa Kundur (University of Toronto)Introduction to Digital Signal Processing

51 / 51

3T

50 / 51