CAN207 Signals and Systems

Part 2 – Discrete-Time Signals and Systems

Lecture-D6
Z-Transform

Zhao Wang
[email protected]
Room EE322

1
Content
• 1. What is z-transform?
– DTFT and z-transform
– Convergence
• 2. Region of Convergence (ROC)
• 3. Inverse DTFT
• 4. Z-transform properties
• 5. Rational Z-transform
• 6. CCLDE and Z-transform
– Zero/pole positions
– Frequency responses

2
1.1 Why do we need another transform?
• Think about all the transforms you have seen so far
– Laplace transform, Fourier series, CTFT, DTFT and DFT
• Why do we need another one?
– Convergence issues with the Fourier transforms:
The DTFT of a sequence exists if and only if the sequence x[n] is
absolutely summable, that is, if

– DTFT may not exist for certain signals of practical interest or some
analytical signals, whose frequency analysis can therefore not be
obtained through DTFT

3
1.2 Z-Transform
• A generalization of the DTFT leads to the z-transform that
may exist for many signals for which the DTFT does not.
– DTFT is in fact a special case of the z-transform
– …just like the CTFT is a special case of Laplace’s transform.
• Importance of z-transform
– The use of z-transform techniques permits simple algebraic
manipulations
– The z-transform has become an important tool in the analysis and
design of digital filters
– The representation of an LTI discrete-time system in the z-domain
is given by its transfer function which is the z-transform of the
impulse response of the system

4
1.2 Z-Transform
• For a given sequence x[n], its z-transform X(z) is defined as
  n 
X[ z]  
n 
x[n]z  n  
n 
 
x[n] re j  
n 
x[ n]r  n e  j n

where z lies in the complex space, that is : z=a+jb=rejω

– It follows that the DTFT is indeed a special case of the z-transform,
specifically, z-transform reduces to DTFT for the special case of
r=1, that is, |z|=1.
– The contour |z|=1 is a circle in the
z-plane of unit radius -> the unit circle
– Hence, the DTFT is really the z-
transform evaluated on the unit circle.
X( )  X[ z ] e j

5
1.2 Convergence
• Just like the DTFT, z-transform also has its own
convergence requirements: x[n]r-n must be absolutely
summable, that is,

• For a given sequence, the set R of values of z for which its

z-transform converges is called the region of convergence
(ROC).
– The area where the above condition is satisfied defines the ROC,
which in general is an annular region of the z-plane
R  z  R where 0  R   R   
– The z-transform must always be specified with its ROC !
DTFT exists only when ROC include |z|=1, the unit circle!
6
Example 1
• Determine the z-transform and the corresponding ROC of the
unit step sequence u[n]
 
U[ z ]   u[n]z  n   z  n  1  z 1  z 2    z  n  
n  n0

which converges to
1
U[ z ]  1
, for z 1  1
1z
z
 , for z  1
z1
• The region of convergence is the annular region in the z-plane
1 z 

7
Example 2
• Determine the z-transform and the corresponding ROC of the
causal sequence x[n]=αnu[n] (right-sided)
  n 1
X[ z ]  
n 
n
 u[n]z n
  z
n0
 1
 X[ z] 
1   z 1
, for  z 1  1

z
 , for   z
z 

8
Example 3 Is the same with that in previous slide, but with different ROC

• Now consider the anti-causal y[n]=-αnu[-n-1] (left-sided)

 1  
n n n n m m 1
Y[ z ]    u[ n  1]z    z   z   z    m zm
n  n  m1 m0

 1 z 1 z
 1
 1
 , for z  
1  z 1 z z 

9
1.3 Impulse response and transfer function
• Impulse responses: x[n]=αnu[n] and y[n]=-αnu[-n-1]
• Transfer functions: X[ z]  z , for z  
z 
z
Y[ z ]  , for z  
z 
– The z-transforms of the two sequences x[n] and y[n] are identical
even though the two parent sequences are different
– Only way a unique sequence can be associated with a z-transform is
by specifying its ROC
– Both transfer functions have a pole at z=α, which make the transfer
function asymptotically approach to infinity at this value. Therefore,
z=α is not included in either of the ROCs.

10
2.1 ROC of the z-transform

[ ]
= = , | |>| |
[ ] −
• In the X[z] given above, z = 0 is its zero, and z = α is its pole.
• The circle with the radius of α is called the pole circle. A system may
have many poles, and hence many pole circles.
• For right sided sequences, the ROCs extend outside of the outermost
pole circle, whereas for left sided sequences, the ROCs are the inside of
the innermost pole circle.
• For two-sided sequences, the ROC will be the intersection of the two
ROC areas corresponding to the left and right sides of the sequence.

11
2.1 ROC of the z-Transform
• For double sided sequence:
= − [− − 1]
• Its z-transform is:
1 1
= +
1− 1−

– Two poles of the transfer function: = and =

– ROC: > and <
Left-sided Right-sided
R2 R1

|β|
|α| |β|

12
2.1 ROC of the z-Transform
• When R1 < R2
Right-sided
Left-sided
R1  z  R2
R2
if 0  R1  R2   R1
|β|

R1
ROC of a left-sided ROC of a right-sided
sequence is inside of sequence is outside of
a circular area a circular area
R2

• When R1 > R2
– No valid ROC => z-transform doesn’t exist.
13
Example 4
• Consider x[n]=5nu[n]-8nu[-n-1]
z z
X[ z]  
z5 z8
– Corresponding ROCs are |z|>5 and |z|<8
– Therefore the ROC for this signal is the annular region 5<|z|<8

• Consider x[n]=8nu[n]-5nu[-n-1]
z z
X[ z ]  
z5 z8
– Corresponding ROCs are |z|<5 and |z|>8
– Therefore, the z-transform of this sequence
does not exist!

14
 2.2 Existence of DTFT and z-transform
• Since DTFT is the z-transform evaluated on the unit circle,
that is for z=ejω, DTFT of a sequence exists if and only if the
ROC includes the unit circle!
– The DTFT for x[n]=5nu[n]-8nu[-n-1] clearly does not exist, since the
ROC does not include the unit circle!

– Consider the sequence x[n]=0.9nu[n]-1.1nu[-n-1]

• Its transfer function is: X[ z ] 
z z

z  0.9 z  1.1
• with the ROC as 0.9<|z|<1.1, which includes the unit circle
• Therefore, the DTFT of x[n] exists

The existence of DTFT is not a guarantee for the existence of the z-transform either!

15
3.1 Inverse z-transform
• The inverse z-transform is defined as

where C is a counter-clockwise contour encircling the origin in the

ROC of X(z) gives the contour integral
• There are three methods for the evaluation of the inverse z-
transform in practice
– 0. Observe the X[z] and directly get x[n] from the commonly used z-
transform pair;
– 1. Direct evaluation by the contour integration using the Cauchy
Residue theorem
– 2. Long division of the numerator by the denominator
– 3. Partial-fraction expansion and table lookup 16
Commonly used z-transform pairs

17
3.2 Inverse z-Transform by long division
• The z-transform of a causal sequence can be expanded in a
power series in z-1.
• For a rational z-transform expressed as a ratio of polynomials
in z-1, the power series expansion can be obtained by long
division.

• Example 6 – Evaluate the inverse z-transform of

Using the long division

Not close-form expression, not good enough!

18
3.3 Inverse z-Transform by Partial Fraction Expansion
• A rational H(z) can be expressed as

• If M ≥ N then H(z) can be re-expressed through long division

where the degree of P1(z) is less than N. The rational fraction

P1(z)/D(z) is then called a proper polynomial .

19
3.3 Inverse z-Transform by Partial Fraction Expansion

• Simple Poles: In most practical cases, the rational z-transform

of interest H(z) is a proper fraction with simple poles, then it
can be written in the following form

is not the inverse transform of the original H(z) we are interested in

• So, we simply compute the partial fraction of H(z)/z, which

will then give us the inverse z-transform of H(z)

20
3.3 Inverse z-Transform by Partial Fraction Expansion

• The constants Ai, which are the residues at the poles of

H(z)/z, can be computed as follows:

• Example 7: Find the inverse z-transform of H(z) given

the ROC
– i) 0.2<|z|<0.6
– ii) |z|>0.6

21
3.3 Inverse z-Transform by Partial Fraction Expansion
• Multiple Poles: If the z-domain function contains an m-multiple pole,
that is, a term as the following is included

• this term is expanded as follows:

where each coefficient can be computed by taking consecutive

derivatives and evaluating the function at the pole

22
4.1 Z-transform properties

• Linearity

– ROC of X(z) is the intersection of ROCs of X1(z) and X2(z)

• Example
– 1. Determine the z-Transform and the ROC of
– 2. Determine the z-Transform and the ROC of
– 3. Determine the z-Transform and the ROC of

23
4.2 Z-transform properties

• Time-shifting

And the ROC remains unchanged except for z = 0 if k > 0 and

z = ∞ if k < 0

• Example
– Determine the z-Transform of the signal

24
4.3 Z-transform properties

• Scaling in the z-domain

For any constant a

• Example
– Determine the z-transform and its ROC of the causal sequence

25
4.4 Z-transform properties

• Time Reversal

• Example
– Determine the z-transform and its ROC of

26
4.5 Z-transform properties

• Differentiation in the z-Domain

ROC remains unchanged

• Example
– Find the z-Transform of

27
4.6 Z-transform properties

• Convolution of Two Sequences

The ROC is the intersection of that for X1(z) and X2(z)

28
4.7 Z-transform properties

• Parseval’s relation

29
4.8 Properties Summary

30
 5.1 Rational z-transform

• The z-transforms of LTI systems can be expressed as a ratio

of two polynomials in z-1, hence they are rational transforms.
– Starting with the constant coefficient linear difference equation
(CCLDE) representation of an LTI system:

Z Z
Y[z]  a1z1Y[z]  a2 z2Y[z]  aN zNY[z]  X[z]  b1z1X[z]  b2 z2 X[z]  bM z MX[z]

Y[ z] b0  b1 z 1  b2 z 2    bM z M A ratio of two polynomials

H[ z]  
X[ z] a0  a1 z 1  a2 z 2    aN z  N Degree of H(z) is max{M,N}
31
5.1 Rational z-transform

• A rational z-transform can be alternately written in factored

form as

– At a root z = ξl of the numerator polynomial, H(ξl) = 0 and these

values of z are called the zeroes of H(z)
– At a root z = pℓ of the denominator polynomial H(pℓ)->∞, and as a
result, these values of z are known as the poles of H(z)
– There are M finite zeroes and N finite poles of H(z)
– There are additional (N-M) zeros at the origin if N>M or (N-M)
poles at z = 0 if N<M

32
5.1 Rational z-transform

• Example: z-Transform of the Unit Step

– The region of convergence in the z-plane

The ROC of a rational z-

transform is bounded by
the locations of its poles

33
• Example: A physical interpretation of the concepts of poles
and zeros can be given by plotting the log-magnitude
20log10|G(z)| of G(z) The poles are at z=0.4±j0.6928
The zeroes are at z=1.2±j1.2

34
clear;
close all;
N=256;
rez=linspace(-4,4,N);
imz=linspace(-4,4,N);
%create a uniform z-plane
for n=1:N
z(n,:)=ones(1,N).*rez(n)+j*ones(1,N).*imz(1:N);
end
%Compute the H function on the z-plane
for n=1:N
for m=1:N
Hz(n,m)=(1-2.4*z(n,m)^(-1)+2.88*z(n,m)^(-
2))/(1-0.8*z(n,m)^(-1)+0.64*z(n,m)^(-2));
end
end
%Logarithmic mesh plot of the H function
mesh(rez, imz, 20*log10(abs(Hz)))

35
• Matlab has simple functions to determine and plot the poles and zeros of
a function in the z-plane
– tf2zpk():[Z,P,K]=tf2zpk(NUM,DEN) finds the zeros, poles, and gain.
b=[1 -2.4 2.88]; z = 1.2000 + 1.2000i
a=[1 -0.8 0.64]; 1.2000 - 1.2000i
[z,p,k] = tf2zpk(b,a) p = 0.4000 + 0.6928i
0.4000 - 0.6928i
k=1
– [num,den] = zp2tf(z,p,k) implements the reverse process
– zplane(): zplane(Z,P) plots the zeros Z and poles P (in column vectors)
with the unit circle for reference.
zplane(B,A) plots the poles and zeros of B(z)/A(z) where B and
A are row vectors containing transfer function polynomial coefficients
zplane(b,a);
zplane(z,p);

36
 5.2 Frequency Response

[H w]=freqz([1 -2.4 2.88],[1 -0.8 0.64],256);

figure
plot(w/pi, abs(H))
grid
title('Transfer function')
xlabel('Frequency \omega / \pi')

This system has two zeros at

z=1.2±j1.2 and two poles at
z=0.4±j0.6928

37
5.3 Poles and Zeros
• The ROC of a rational z-transform cannot
contain any poles and is bounded by the poles
– For a right sided sequence, the ROC is outside of
the largest pole
– For a left sided sequence, the ROC is inside of the
smallest pole
– For a two sided sequence, some of the poles
contribute to terms in the parent sequence for n<0
and other to terms for n>0. Therefore, the ROC is
between two circular regions: outside of the largest
pole coming from the n>0 sequence and inside of
the smallest pole coming from the n<0 sequence.
– If the sequence is of finite length, then the ROC
includes the entire z-plane, except possibly z=0
and/or z=∞.
38
5.3 Stability & ROC
– Now, for an LTI system to be stable it must be absolutely summable,
or in other words, it must have a DTFT. But for a system to have a
DTFT, its ROC must include the unit circle.
• An LTI system is stable, if and only if the ROC of its transfer
function H(z) includes the unit circle!

– Furthermore, a causal system’s ROC lies outside of a pole circle. If

that system is also stable, its ROC must include unit circle
• Then a causal system is stable, if and only if, all poles are
inside the unit circle!
– Similarly, an anti-causal system is stable, if and only if its poles lie
outside the unit circle.

39
Behavior of a Single Real-Pole Causal Signal

40
Behavior of a Double Real-Pole Causal Signal

41
Behavior of a Causal Signal with a Pair of
Complex-Conjugate Poles

42
5. Summary

• Analysing the transfer function:

– Representation: polynomial ratio VS factor form
– Roots of numerator  zeroes  H(zero) = 0
– Roots of denominator  poles  H(pole) ∞
• Pole circles: bounds the ROC

• For an LTI system to be causal and stable

– ROC includes |z|=1
All poles inside the unit circle
– Right sided

43
6.1 CCLDE coefficients
• All discrete systems can be represented using Constant Coefficient,
Linear Difference Equations (CCLDE), of the form

• The function H(z), which is the z-transform of the impulse response h[n]
of the LTI system, is called the transfer function
– Using the CCLDE coefficients

44
6.2 Frequency response and the transfer function
• If the ROC of the transfer function H(z) includes the unit
circle, then the frequency response H(ω) of the LTI digital
filter can be obtained simply as follows:
H    H  e j
  H  z z  e j

• So the frequency response of a typical LTI system is

• From which we can obtain the magnitude and phase response

45
6.2 Frequency response and the transfer function

– The magnitude response |H(ω)| at a specific value of ω is given by

the product of the distances to all zeros divided by the product of the
distances to all poles!

– The phase response at a specific value of ω is obtained by adding the

phase of the term b0/a0 and the linear-phase term ω(N-M) to the sum
of the angles of the zero vectors minus the angles of the pole vectors

46
6.3 Frequency response by pole and zero distances

• Example:
– The transfer function of a filter has zeros at zo  ro e  jo and poles
 j p
z 
at p p r e , thus
( z  ro e jo )( z  ro e  jo ) z 2  2 ro cos o z  ro2
H ( z)  j p  j p
 2 2
( z  rp e )( z  rp e ) z  2 rp cos  p z  rp

– Choosing ro= 1.2, θo = 300, rp= 0.9, θp = 600

z 2  2.078 z  1.440
H ( z)  2
z  0.900 z  0.810
– The frequency response at ω is given by
j j0 j  j0 
( e  r e )( e  r e ) u u
H ( e j )  j o j p j o  j p  1 2
( e  rp e )( e  rp e ) u3u4

• Where the uk are complex phasors pointing from a zero or pole to the point
ejω on the unit circle.

47
6.3 Frequency response by pole and zero distances
• Thus we may evaluate the frequency response at a given
 frequency in
terms of the magnitudes and angles of the phasors uk
product of distances to zeros
H ( e j ) 
product of distances to poles

Arg{H ( e j )}  (sum of angles from zeros to e j )

-(sum of angles from poles to e j )

This pole – zero diagram shows u1

u2 u3 and u4 for our example

48
– An approximate plot of the magnitude and phase responses of the
transfer function of an LTI digital filter can be developed by
examining the pole and zero locations
– Now, the frequency response has the smallest magnitude around
ω=ζ, and the largest magnitude around ω=p.
– Of course, at ω=p, the response is infinitely large, and at ω=ζ, the
response is zero
6 0

-1
5

-2
4
magnitude response

-3

phase
3
-4

2
-5

1
-6

0 -7
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
frequency frequency

49
6.3 Frequency response by pole and zero distances

• Therefore:
– To highly attenuate signal components in a specified frequency
range, we need to place zeros very close to or on the unit circle in
this range.

– Likewise, to highly emphasize

signal components in a
specified frequency range, we
need to place poles very close
to or on the unit circle in this
range.

50
6.4 Graphical interpretation
• Complex vector
– αk is in general a complex quantity, let’s write that as  k   k e j k

– Then we have e j   k  e j   k e jk

– the term e j   k e j represents a vector in the z-plane, that starts at

k

the point z   k e j and ends at the point z  e j , which is on the

k

unit circle
– As ω varies from 0 to 2π, the tip of this vector moves counter-
clockwise tracing the unit circle.

Zero vectors: αk is ξk, in numerator;

Pole vectors: αk is pk, in denominator.

51
• The magnitude response |H(ω)|, at a given frequency ω, is the
product of the magnitude (length of orange vector) of all zeros,
divided by the magnitude of all poles, as evaluated at that ω.
– If αk is a zero (i.e., a numerator factor), the
overall magnitude vector of H(ω) will be
small at frequencies around φk, and will be
exactly zero if αk is on the unit circle, causing
H(ωk) = 0.
– Conversely, if αk is a pole (i.e., a denominator
factor), the overall magnitude vector of H(ω)
will be large at frequencies around φk, and
will go to infinity if αk is on the unit circle.
– This is why the zeros and the poles that are at
or close to the unit circle have a larger impact
on the overall frequency response than those
that are further away from the unit circle.

52
6.4 Graphical interpretation – An example

One zero at z = 0
One pair of conjugate poles
at z = -0.5000 ± j0.5000

53
6.4 Graphical interpretation – An example
• Consider the M-point moving-average FIR filter with an
impulse response

– The transfer function has M zeros on

the unit circle at
– There are M-1 poles at z = 0 and
a single pole at z = 1
– The pole at z = 1 exactly cancels the
zero at z = 1
– The ROC is the entire z-plane except
z=0 54
Moving Average Filter

55
6. Summary

• Frequency response
– Magnitude response Z-transform
– Phase response (Zero-pole positions)

• Graphic explanation
• Filter design based on zero position arrangement

56
Lect-D6 Practices
• Exercise 1. Consider the z-transform X(z) whose pole-zero plot is
shown on the right.
– A) Determine the ROC of X(z)
if it is known that the Fourier
transform exists. For this case,
determine whether the corres-
ponding sequence x[n] is right
sided, left sided, or two sided.

– B) How many possible two-

sided sequences have the pole-zero plot as shown?

– C) Is it possible for the pole-zero plot as shown to be associated with a sequence

that is both stable and causal? If so, give the appropriate region of convergence.

57
Lect-D6 Practices
• Exercise 2.
– a) What are the conditions for a causal LTI system to be stable?
.
– b) Is the system H z = stable?
.
– c) Find the ROC of a stable H(z).

• Exercise 3. Determine the ROC for each of the following sums to

converge:
– a) ∑

– b) ∑
( )
– c) ∑

58
Lect-D6 Practices
• Exercise 4. The input to a causal LTI system is

• the z-transform of the output of this system is

– A) Determine H(z), the z-transform of the system impulse response.

Be sure to specify the ROC.

– B) What is the ROC of Y(z)?

– C) Determine y[n].
59
Lect-D6 Practices
• Exercise 5. Consider the causal sequence x n = (−0.7) , with a
z-transform given by X(z). Determine the inverse z-transform of X(z3)
without computing X(z).

• Exercise 6. Sketch roughly the magnitude response ( ) of the

Fourier transforms corresponding to the pole-zero patterns given below:

where circle represents the zeros and cross represents the poles.
60