Integral Transforms

• In mathematics and physics we frequently

encounter pairs of functions related by an
expression of the form:
b
g ( )  
a
f ( t ) k ( , t ) dt

This expression is called an integral transform

where K(α,t) is the kernel of transformation

1 Dr. B. Khan CECOS University

 Unlimited integral transforms
• The types of integral transform are numerous,
for example Laplace transform, Fourier
Transform, Hankel transform , Mellin transform
, Borel transform ect.
• Laplace transform is one of these many
transforms and is heavily used in physics and
engineering for the analysis of LT systems such
as electric circuits, harmonic oscillators, optical
devices , and mechanical systems
2 Dr. B. Khan CECOS Unversity
 ?What these transformations do
• These transformations are interpreted as transformations
from time-domain in which inputs and outputs are
functions of time, to the domain where the same inputs
and outputs are functions of complex angular frequency
in radians per unit time.
• Given a simple mathematical or functional description of
an input to or output from a system, these transforms
provide an alternative functional description that often
simplifies the process of analyzing the behavior of the
system or synthesizing a new system based on a set of
specifications.
3 Dr. B. Khan CECOS University
 z-transform
• The z-transform is for discrete-signals and is
defined as the power series of the form:

X (z)  
x(n) z n

where z is complex variable.

• Since the z-transform is an infinite power series,
it exists only for those values of z for which the
series converges. Such a set of values is called
the region of convergence (ROC) of X(z).
4 Dr. B. Khan CECOS University
 Inverse z-transform
• The inverse z-transform is defined as:
1

n 1
x(n)  X ( z ) z dz
2j C
where c is a closed contour within the ROC of X(z).
• Note that the z-transformation and its ROC go to
gather.

5 Dr. B. Khan CECOS University

 Definition of ROC
• Given a sequence, the set of values of z for
which the z-transform converges, i.e., |X(z)|
<, is called the region of convergence.
 
| X ( z ) |  x (
n  
n ) z n
  | x
n  
( n ) || z | n


ROC
ROC isis centered
centered at at origin
origin and
and
consists
consists of
of aa set
set of
of rings.
rings.
6 Dr. B. Khan CECOS University
 Region of convergence(ROC)
• Let z = rejθ so that we can write
 
| X ( z ) |  x ( n) r
n  
n
e  jn
  x (n) r
n  
n
e  jn


| X ( z ) | 
n  
x(n)r n

1 
x(n)
| X ( z ) | 
n  
x(n)r n
 
n0 r n

7 Dr. B. Khan CECOS University

 Region of convergence(ROC)
• Or  
x(n)
| X ( z ) | 
n 1
x( n)r n
 
n0 r n

• If the first sum converges, there exit values of r

small enough such that the product sequence
x(-n)rn, 1 ≤ n ≤ ∞ is absolutely summable.

• Therefore, the ROC for the first sum consists

of all points in a circle of some radius r1, where
r1< ∞.
8 Dr. B. Khan CECOS University
 Region of convergence(ROC)
• For the second sum to converge, there must exist
values of r large enough such that the product
sequence x(n)/rn, 0≤ n < ∞, is absolutely
summable.
• The ROC of X(z) is therefore, generally the annular
region in the z-plane, r2 <r <r1 where both sums
are finite as shown in the figure.
• If r2 > r1, then there is no common region where
both sums are finite and hence X(z) does not exist.

9 Dr. B. Khan CECOS University

 Region of Convergence ROC
 
| X ( z ) |  x (
n  
n ) z n
  | x
n  
( n ) || z | n


Im
ROC
ROC isis an
an annular
annular ring
ring centered
centered atat the
the
origin.
origin.

Re
ROC  r2  | z | r1

10 Dr. B. Khan CECOS University

 Example: ROC of
a right sided sequence

x(n)  a u (n) n

x(n)

... n

8- 7 - 6 - 5 - 4 - 3 - 2- 1- 1 2 3 4 5 6 7 8 9 10

11 Dr. B. Khan CECOS University

 ROC of a causal sequence

x(n)  a nu (n) For convergence of X(z), we require that



 |
| az 1
| az 1
| 1
X ( z)   a u (n)z
n  
n n
n 0

| z || a |

  a n z n 
1 z
n 0 X ( z )   (az 1 ) n  
n0 1  az 1 z  a

  (az 1 ) n | z || a |
n0
12 Dr. B. Khan CECOS University
 A a causal sequence
ROC for x(n)=anu(n)

z
X ( z)  , | z || a | ??Which
za Which one
oneImisis stable
stable
Im

1
1 a a
Re
a a
Re

13 Dr. B. Khan CECOS University

 A anti-causal sequence

x(n)   a u (n  1)
n 1
X ( z)  1
, | z || a |
1  az
z
Im X ( z)  , | z || a |
za

ROC is bounded by the

Re pole and is the interior
a
of a circle.

14 Dr. B. Khan CECOS University

 Region of convergence(ROC)
• The causal and anti-causal sequences have the
same closed-form expression indicating that z-
transform alone does not uniquely specify the
signal in the time-domain.
• However, the z-transform for the two sequences
has different regions of convergence.
• The ambiguity is therefore resolved if the ROC
of X(z) is also specified.

15 Dr. B. Khan CECOS University

 A non-causal sequence

x(n)  ( 13 ) n u (n)  ( 12 ) n u (n  1)

z z 2 z ( z  121 )
X ( z)   
z3 z2
1 1 ( z  13 )( z  12 )
Im

ROC is bounded by poles and

is a ring.
1/1
1/3 2 1/2 Re

ROC does not include any pole.

16 Dr. B. Khan CECOS University

 Region of convergence(ROC)
• The ROC also depends on the duration of the
sequence or discrete-time signal.
• The ROC of infinite duration causal signal is the
exterior of circle with the radius equal to the
outermost pole.
• The ROC of infinite duration anti-causal signal is
the interior of circle with the radius equal to the
innermost pole.
• The ROC of two-sided infinite duration signal is a
ring (annular region) in the z-plane.
17 Dr. B. Khan CECOS University
 The ROC of finite-duration sequences
• The series is finite as long as each of the terms:
n
x(n) z 
• In general it may not be possible or may be
unnecessary to express the sum of finite set of
terms in closed form.
• Let us consider the convergence problem of
two-sided, right-sided and left-sided finite
discrete-time signals.

18 Dr. B. Khan CECOS University

 The ROC of finite-duration sequences
• The ROC also depends on the duration of the
sequence or discrete-time signal.
• A finite- duration sequence is defined as the
sequence which has a finite number of nonzero
values in the interval say n= N1 to n=N2 where N1
and N2 are finite and may be positive or
negative.
• A right-sided sequence is defined as
x ( n )  0, n  N 1
19 Dr. B. Khan CECOS University
 The ROC of finite-duration sequences
• The z-transform of right-sided sequence is,
therefore 
X (z)   x(n) z
n  N1
n

1 
X ( z)   x ( n) z
n  N1
n
  x ( n) z
n 0
n

N1 
X ( z)   x(n) z   x(n) z n n

n 1 n 0

20 Dr. B. Khan CECOS University

 The ROC of finite-duration sequences
• The first sum converges for all z except at z = ∞
• The second sum converges for all z  r2

• The ROC is therefore r2  z   for N1<0

• The ROC is z  r2 for N1>0

21 Dr. B. Khan CECOS University

 The ROC of finite-duration sequences
• A left-sided sequence is defined as
x ( n )  0, n  N 2
• The z-transform of a left-sided sequence is
N2
X (z)   x(n) z
n  
n

0 N2
X ( z)   x(n) z
n  
n
  x (n) z
n 1
n

22 Dr. B. Khan CECOS University

 The ROC of finite-duration sequences
• Or
 N2
X ( z )   x( n) z   x (n) z n n

n 0 n 1
• The first sum converges for some z  r1
• The second sum converges for all z except at
z= 0.
• ROC of X(z) is 0  z  r , N  0
1 2

z  r1 , N 2  0
23 Dr. B. Khan CECOS University
 The ROC of finite-duration sequences
• A finite- duration two-sided sequence is defined
as the sequence which has a finite number of
nonzero values in the interval say n= N1 to n=N2
where N1 and N2 are finite and may be positive
or negative. It is defined as:
x ( n )  0, N 2  n  N 1
• The z-transform of two-sided finite sequence is
N 2

X (z)  
n N1
x(n ) z n

24 Dr. B. Khan CECOS University

 The ROC of finite-duration sequences
• The z-transform of right-sided sequence is,
therefore
1 N2
X ( z)   x ( n) z
n  N1
n
  x ( n) z
n 0
n

N1 N2
X ( z)   x(n) z   x(n) z n n

n 1 n 0

25 Dr. B. Khan CECOS University

 The ROC of finite-duration sequences
• The first sum converges every where except at
z=∞
• The second sum converges for all z except at
z= 0.
• ROC of X(z) is 0 z 

Example 3.1.1 PRO

26 Dr. B. Khan CECOS University

 Properties of ROC
• A ring or disk in the z-plane centered at the origin.

• The ROC cannot include any poles.

• Finite Duration Sequences: The ROC is the entire z-plane except possibly at z = 0
(Right-sided) or z =  (left-sided).

• Two-sided Finite Duration Sequences : The ROC is the entire z-plane except
possibly at z = 0 and z = .

• Right sided sequences: The ROC extends outward from the outermost finite pole
in X(z) to z = .

• Left sided sequences: The ROC extends inward from the innermost nonzero pole
in X(z) to z=0.

27 Dr. B. Khan CECOS University

 Properties of ROC continued
• For rational X(z): The ROC is bounded by poles or extend to
infinity.

• For rational X(z): If x(n) is right-sided, then the ROC is the

region in the Z-plane outside the outermost pole, that is outside
the circle of radius equal to the largest magnitude of the poles
of X(z).

• For rational X(z): If x(n) is left-sided, then the ROC is the

region in the Z-plane inside the innermost nonzero pole, that is,
inside the circle of radius equal to the smallest magnitude of the
poles of X(z)
28 Dr. B. Khan CECOS University
 Rational z-transform
• A z-transform which is in the form of a ratio of two
polynomials in z-1 (or z) is called rational z-transform.
Mathematically B (z)
X (z) 
A(z)
M


k 0
bk z k
X (z)  N


k 0
a k z k

29 Dr. B. Khan CECOS University

 Factorized form of X(z)
• The roots of the denominator (pk) are called the
poles and the roots of the numerator (zk) are
called the zeros of X(Z) M

N M
 (z  zk )
X ( z )  Gz k 1
N


k 1
(z  pk )
• where G  b0 / a0

30 Dr. B. Khan CECOS University

 Partial Expansion of X(z)
• If X(z) a proper rational function, then it can
be expressed as sum of simple fractions by
method of partial fraction expansion. We can
write: A(z)
X (z) 
B(z)
• The partial expansion of X(z) depends on the
nature of the roots of the polynomial B(z).
• The roots may be real or complex, which
might be distinct or repeated.
Dr. B. Khan CECOS University
31
 Partial expansion of X(z)
• Most of the problems in which partial expansion
is required is off intermediate level.
• For complicated partial expansion, a formula is
briefly discussed.
• Suppose that the denominator polynomial has
n1 roots equal to p1, e.i., p1 roots of multiplicity
n1, p2 roots of multiplicity n2, and in general, pr
roots of multiplicity nr such that N=n1+n2+…+nr.

Dr. B. Khan CECOS University

32
 Partial expansion of X(z)
• The partial expansion of such a function is
ni
X (z) r
C ik
z
 
i 1 k 1 (z  pi ) k

where r is the pole number. For example for r

= 1, we have p1 and n1 that is p1 is repeated n1
times.
• The constants Cik are called the residues.

33 Dr. B. Khan CECOS University

 Calculation of residues cik
• The formula for the residue of the ith root of
multiplicity k in the partial expansion, denoted
by cik is:

1 d  ni X ( z ) 
cik  ni  k ( z  pi )  z  pi
(ni  k )! dz  z 

34 Dr. B. Khan CECOS University

 Calculation of the residues
• Like the poles, the residues may be real or
complex.
• There are simple procedures (partial fraction)
as mentioned in the intermediate course for
the calculation of these residues.
• However, there is one universal formula, as
mentioned before, for finding out these
residues

35 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• The characteristic behavior of causal signals
depends on whether the poles of the transform
are inside the unit circle or outside the unit
circle, or on the unit circle.
• The discussion is limited to real causal signals.
1
x(n)  a u(n)  X ( z) 
n
1
, ROC: z  a .
1  az

• X(z) has one pole at p1 = a, and one zero at z1=0.

36 Dr. B. Khan CECOS University
 Pole locations and time-domain
behavior of causal signals
• If the pole is inside the unit circle.

• x(n) decays and is bounded.

37 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is inside the unit circle.

• x(n) again decays and is bounded.

38 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is on the unit circle.

• x(n) remains fixed and is bounded.

39 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is on the unit circle.

• X(n) alternates and is unbounded.

40 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is outside the unit circle.

• X(n) grows and is unbounded.

41 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is outside the unit circle.

• x(n) grows and is unbounded.

42 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• Let x(n) = n anu(n) is causal real signal. The z-
transform of x(n) is:
1
az
X ( z)  , ROC : z  a.
1  az  1 2

• X(z) has a pole of multiplicity 2 at p1 = a, and

one zero at z1=0

43 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is of multiplicity 2 and is inside unit circle.

• x(n) grows, then decays but remains unbounded.

44 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is of multiplicity 2 and is inside unit
circle.

• x(n) changes sign. It is unbounded.

45 Dr. B. Khan CECOS University
 Pole locations and time-domain
behavior of causal signals
• If the pole is of multiplicity 2 and is on the unit
circle.

• x(n) grows and is unbounded.

46 Dr. B. Khan CECOS University
 Pole locations and time-domain
behavior of causal signals
• If the pole is of multiplicity 2 and is on the unit
circle.

• x(n) grows and is unbounded.

47 Dr. B. Khan CECOS University
 Pole locations and time-domain
behavior of causal signals
• If the pole is of multiplicity 2 and outside the
unit circle.

• x(n) grows and is unbounded.

48 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is of multiplicity 2 and outside the
unit circle.

• x(n) changes sign. It is unbounded.

49 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is complex and inside the unit circle.

• x(n) oscillates, decays and is bounded.

50 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is complex and is on the unit circle.

• x(n) oscillates, remains fixed and is bounded.

51 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• If the pole is complex and outside the unit circle.

• x(n) oscillates , grows and is unbounded.

52 Dr. B. Khan CECOS University

 Pole locations and time-domain
behavior of causal signals
• The effects of the location of a single are
summarized as:
• Causal real signals with simple real poles or
simple complex conjugate pairs, which are
inside or on the unit circle, are always bounded.
• A signal with a pole (or complex-conjugate pair
of poles) near the origin decays more rapidly
than one associated with a pole near (but
inside) the unit circle.
53 Dr. B. Khan CECOS University
 The system function of an LTI system
• From the convolution, the output of a relaxed LTI
system is y(n) = x(n)*h(n), where h(n) is the unit
sample response of the system. Convolution
property of z-transform:
Y (z)
Y (z)  X (z)H (z)  H (z) 

X (z)
H ( z)   h(n) z
n  
n

• H(z) is the z-domain characterization of the

system and is called the system function.
54 Dr. B. Khan CECOS University
 The system function of an LTI system
• An LTI system can be described by a linear
difference equation with constant coefficients of
the form, assuming that a0 = 1:
N M
y (n )    a k y (n  k )   bk x (n  k )
k 1 k 0

N M
Y ( z )   a k Y ( z ) z k
  bk X ( z ) z k

k 1 k 0

55 Dr. B. Khan CECOS University

 The system function of an LTI system
• Rearranging and simplifying we get
 N
 M
Y ( z ) 1 

a
k 1
k z k
  X ( z ) bk z
 k 0
k

Y (z)
 bk z k

 H (z)  k 0
N
X (z)
1  k 1
ak z k

56 Dr. B. Khan CECOS University