Discrete-Time Signal processing

the Z-transform

2020/12/30
 Chapter 3 The z-Transform

3.0 Introduction
3.1 z-Transform
3.2 Properties of the Region of
Convergence for the z-transform
3.3 The inverse z-Transform
3.4 z-Transform Properties

2 2020/12/30
 3.0 Introduction

Fourier transform plays a key role in analyzing

and representing discrete-time signals and
systems, but does not converge for all signals.
Continuous systems: Laplace transform is a
generalization of the Fourier transform.
Discrete systems : z-transform, generalization
of DTFT, converges for a broader class of
signals.
3 2020/12/30
 3.0 Introduction
Motivation of z-transform:
The Fourier transform does not
converge for all sequences and it is
useful to have a generalization of the
Fourier transform.
In analytical problems the z-Transform
notation is more convenient than the
Fourier transform notation.

4 2020/12/30
 3.1 z-Transform
z-Transform: two-sided, bilateral z-transform

X  z   x  n z
n 
n
Z  x[n]
Z
x[n]  X  z 
one-sided, unilateral z-transform

X z    xnz n

n 0
If z e jw
, z-transform is Fourier transform.

X e jw
   x  n e  jwn
n 
5 2020/12/30
Relationship between z-transform and
Fourier transform
Express the complex variable z in polar
form as
z  re
jw

    xnr e

jw n  jwn
X re
n  

The Fourier transform of the product of xn

n
and the exponential sequence r
 X e
If r  1, X  Z   jw

6 2020/12/30
 Complex z plane
  w    unit circle

7 2020/12/30
 Region of convergence (ROC)
For any given sequence, the set of values of z
for which the z-transform converges is called
the Region Of Convergence (ROC).
z  re jw

    xnr e
 
X z   x  n z  n X re jw n  jwn
n 
n  

Absolute X  re jw
   x n r n

Summability n 

X  z    x  n z
n

n 0
the ROC consists of all values of z such that the
inequality in the above holds
8 2020/12/30
 Region of convergence (ROC)

X  z    x  n z
n

n 0
Convergence of the z-transform for a given
sequence depends only on r  z .
if some value of z, say,
z =z1, is in the ROC,
then all values of z on the
circle defined by |z|=|z1|
will also be in the ROC. z1
if ROC includes unit circle, then Fourier transform and all its
derivatives with respect to w must be continuous functions of w.
 Zero and pole
The z-transform is most useful when the
infinite sum can be expressed in closed form,
usually a ratio of polynomials in z (or z-1).

Pz 
X z  
Qz 

Zero: The value of z for which X z   0

Pole: The value of z for which X z   

10 2020/12/30
 Example 3.1:
Right-sided exponential sequence
Determine the z-transform, including the ROC
in z-plane and a sketch of the pole-zero-plot,
for sequence:
xn a u n
n
 
Solution:   1 z
X  z    a z    az  
n n 1 n
1

n 0 n 0 1  az za
1
ROC: az  1 or z  a

zero : z  0 pole : z  a
11 2020/12/30
 xn  a un
n

z
X  z 
za
: zeros
 : poles for z  a
Gray region: ROC
12 2020/12/30
 Ex. 3.2 Left-sided exponential sequence
Determine the z-transform, including the ROC,
pole-zero-plot, for sequence:
xn  a u n 1
n

Solution:  1
X  z     a u  n  1 z
n n
 a z n n

n  n 
  1
  a z    a z    a z z
n n 1 n

1

n 1 n 1 1 a z z  a
ROC: z  a, zero : z  0 pole : z  a
13 2020/12/30
 xn  a u n 1
n

z
X  z 
za

for z  a

14 2020/12/30
Ex. 3.3 Sum of two exponential sequences
Determine the z-transform, including the ROC,
pole-zero-plot, for sequence:
n n
1  1
x  n    u  n     u  n
Solution: 2  3
  1 n  1
n
  n
X  z      u  n      u  n  z
n   2   3 
 n  n
1  1
    u  n z      u  n z  n
n

n   2  n   3
 n  n
 1 1   1 1 
  z      z 
15 n 0  2
2020/12/30  n   3 
 Example 3.3:
Sum of two exponential sequences
 n  n
 1 1   1 1 
X  z   z      z 
n 0  2  n   3 
 1
1 1 2 z  z  
   12 
1 1 1 1   1  1 
1 z 1 z 1  z 1
1  z 1
2 3   
 2  3 
1 1 1
ROC: z  and z   ROC : z 
2 3 2
16 2020/12/30
 1
1 1
1 z
2

1
1 1  1
1 z 2z  z  
3  12 
 1 1  1 1 
1  z  1  z 
 2  3 
n n
1  1
x  n    u  n     u  n
2  3
17 2020/12/30
 Example 3.4: Sum of two exponential

 
n n
1  1 xn a u n
n
x  n    u  n     u  n
2  3
z
Solution:
n  
X z 
1 Z 1 1 z  a
  un , z
2 1 1 2 for z  a
1 z
n 2
 1 Z 1 1
   un , z
 3 1 1 3
1 z
3
n n
1  1 Z 1 1 1
  un     un  , z
2  3 1 1 1 1 2
1 z 1 z
18 2020/12/30 2 3 ROC:
 1
1 1
1 z
2

1
1 1  1
1 z 2z  z  
3  12 
 1 1  1 1 
1  z  1  z 
 2  3 
n n
1  1
x  n    u  n     u  n
2  3
19 2020/12/30
 Example 3.5:
Two-sided exponential sequence
xn  a u n 1
n n n
 1 1
xn     un    u n  1
 3 2 z
Solution: n X  z 
 1 Z
   un
1
, z
1 za
 3 1 1 3
1 z
3
for z  a
n
1 Z 1 1
   u n  1 , z
2 1 1 2
1 z
2  1
2 z z  
1 1  12  1 1
X z     ROC :  z 
1 1 1 1  1 1  1 1  3 2
1 z 1 z 1  z 1  z 
3 2  3  2 
20 2020/12/30
  1
2 z z  
1 1  12  1 1
X z     ROC :  z 
1 1 1 1  1 1  1 1  3 2
1 z 1 z 1  z 1  z 
3 2  3  2 
n n
 1 1
xn     un    u n  1
 3 2

ROC, pole-zero-plot

21 2020/12/30
 Finite-length sequence
N2
X  z   x  n z
n  N1
n

Example :

xn   n   n  5

X z   1  z 5
ROC : z  0

22 2020/12/30
 Example 3.6: Finite-length sequence
Determine the z-transform, including the ROC,
pole-zero-plot, for sequence:
a n , 0  n  N  1
xn  
 0, otherwise
Solution: N 1 N 1
X  z  a z n n
   az 
1 n

n 0 n 0

1   az 
1 N
1 z a N N
  N 1
1  az 1
z za ROC : z  0
23 2020/12/30
 N=16, a is real

1 z a N N
X  z   N 1 ROC : z  0
z za
pole-zero-plot

24 2020/12/30
 z-transform pairs
 n  1, ROC : all z
1
u n  1
, ROC : z  1
1 z
1
 u  n  1  1
, ROC : z  1
1 z
  n  m  z m
,
ROC : all z except 0  if m  0  or   if m  0 

25 2020/12/30
 z-transform pairs
1
a u n 
n
1
, ROC : z  a
1  az
1
 a u  n  1 
n
1
, ROC : z  a
1  az
1
az
na u n 
n
, ROC : z  a
1  az 1 2

1
az
 na u  n  1 
n
, ROC : z  a
1  az  1 2

26 2020/12/30
 z-transform pairs
1  cos w0 z 1
cos w0 nu n  , ROC : z  1
1  2cos w0 z  z
1 2

cos w0 n u  n 
1
2
e jw0n
e
 jw0n
 u n
1 1 1 
 jw0 1
  jw0 1 
2  1 e z 1 e z 

sin w0 nu n 

sin w0 z 1 , ROC : z  1
1  2cos w0 z  z
1 2

27 2020/12/30
 z-transform pairs
1 r cos w 
r n cos w0nu n  1  2r cos w z 01  r 2 z 2 ,
 z 1
ROC : z  r
0

2
r e 
1 n jw0n n  jw0n
 r cos w0 n  u  n   r e u n 
1 1 1 
 jw0 1
  jw0 1 
2  1  re z 1  re z 

r n

sin w0 n u n 
r sin w0 z 1 , ROC : z  r
1  2r cos w0 z  r z
1 2 2

28 2020/12/30
 z-transform pairs

a , 0  n  N  1
n
1 a z N N

  1
,
 0 , otherwise 1  az
ROC : z  0

29 2020/12/30
 3.2 Properties of the ROC for the
z-transform
Property 1: The ROC is a ring or disk in
the z-plane centered at the origin.
0  rR  z  rL  
For a given x[n], ROC is
dependent only on z .
rR u  n   rL a u  n  1
n n

1 1
1
 1
, ROC : rR  z  rL
1  rR z 1  rL z
 3.2 Properties of the ROC for the
z-transform
Property 2: The Fourier transform of xn
converges absolutely if the ROC of the
z-transform of xn includes the unit
circle.
The z-transform reduces to the Fourier
transform when z  1 ie. z  e jw
 
X z    xnz n
X e jw
   x  n e  jwn

n 0 n 

31 2020/12/30
 3.2 Properties of the ROC for the
z-transform

Property 3: The ROC cannot contain

any poles.

 X z  is infinite at a pole and therefore

does not converge.

P z
X  z    x  n z n

n 0 Q z

32 2020/12/30
 3.2 Properties of the ROC for the
z-transform

Property 4: If x n is a finite-duration
sequence, i.e., a sequence that is zero
except in a finite interval : N1  n  N2

 then the ROC is the entire z-plane,

except possible z  0 or z  
N2
X  z  a z
n  N1
n n

33 2020/12/30
 3.2 Properties of the ROC for the z-transform
Property 5: If xnis a right-sided sequence,
i.e., a sequence that is zero for n  N1   ,
the ROC extends outward from the outermost
finite pole in X z  to (may including) z  
N
Proof: x n    Ak  dk  ,n  N ,
n
1
k 1

N1 N1

if r  d N  d N 1   d1 , i.e. r  d N
34 2020/12/30
3.2 Properties of the ROC for the z-transform
Property 6: If xn is a left-sided sequence,
i.e., a sequence that is zero for n  N 2  ,
the ROC extends inward from the innermost
nonzero pole in X z  to z  0 .
N
   Ak  dk  , n  N ,
Proof: x n 
n
2
k 1
n  N2 n  N2

n   n  
r  dk ,
35
r  d1 ,
2020/12/30
, r  dN , r  d1
Zhongguo Liu_Biomedical Engineering_Shandong Univ.
 3.2 Properties of the ROC for the
z-transform
Property 7: A two-sided sequence is
an infinite-duration sequence that is
neither right-sided nor left-sided.
If x  n  is a two-sided sequence, the
ROC will consist of a ring in the z-
plane, bounded on the interior and
exterior by a pole and not containing
any poles.

36 2020/12/30
 3.2 Properties of the ROC for the
z-transform
Property 8: ROC must be a connected region.
for finite-duration sequence
ROC: 0  z   possible
z  0
z 
for right-sided sequence
ROC: rR  z   possible z 
for left-sided sequence
ROC: 0  z  rL possible z 0
for two-sided sequence rR  z  rL
37 2020/12/30
 Example: Different possibilities of
the ROC define different sequences

A system with three poles

38 2020/12/30
 Different possibilities of the ROC.

(b) ROC to a (c) ROC to a

right-sided sequence left-handed sequence
39 2020/12/30
 Unit-circle
included

(d) ROC to a (e) ROC to another

two-sided sequence. two-sided sequence
40 2020/12/30
 LTI system Stability, Causality, and ROC
A z-transform does not uniquely determine
a sequence without specifying the ROC
It’s convenient to specify the ROC implicitly
through time-domain property of a sequence
Consider a LTI system with impulse response
h[n]. The z-transform of h[n] is called the
system function H (z) of the LTI system.
stable system(h[n] is absolutely summable
and therefore has a Fourier transform): ROC
include unit-circle.
causal system (h[n]=0,for n<0) : right sided
41 2020/12/30
 Ex. 3.7 Stability, Causality, and the ROC
Consider a LTI system with impulse
response h[n]. The z-transform of h[n] i.e.
the system function H (z) has the pole-zero
plot shown in Figure. Determine the ROC,
if the system is:
(1) stable system:
(ROC include unit-circle)

(2) causal system:

(right sided sequence)
42 2020/12/30 Zhongguo Liu_Biomedical Engineering_Shandong Univ.
 Ex. 3.7 Stability, Causality, and the ROC
Solution: (1) stable system (ROC include unit-circle),
1
ROC:  z  2 , the impulse response is
2
two-sided, system is non-causal. stable.

43 2020/12/30
 Ex. 3.7 Stability, Causality, and the ROC

(2) causal system: (right sided sequence)

ROC: z  2 ,the impulse response is right-

sided. system is causal but unstable.

 A system is causal
and stable if all the 2
1
2
poles are inside the
unit circle.

44 2020/12/30
 Ex. 3.7 Stability, Causality, and the ROC
1
ROC: z  , the impulse response is
2
left-sided, system is non-causal, unstable
since the ROC does not include unit circle.

45 2020/12/30
 3.3 The Inverse Z-Transform
Formal inverse z-transform is based
on a Cauchy integral theorem.
1

n 1
xn  X ( z ) z dz c  ROC
2 j c
j Im[ z ]

rR rL
0 Re[ z ]

46 2020/12/30
 3.3 The Inverse Z-Transform

Less formal ways are sufficient

and preferable in finding the
inverse z-transform. :
Inspection method
Partial fraction expansion
Power series expansion

47 2020/12/30
 3.3 The inverse z-Transform
3.3.1 Inspection Method
Z1
a u  n 
n
1
, z a
1  az
1 1
X z   , z
1 1 2
1 z
2
n
1
xn    un
2
48 2020/12/30
 3.3 The inverse z-Transform
3.3.1 Inspection Method
Z1
a u  n  1 
n
1
, z a
1  az
1 1
X z   , z
1 1 2
1 z
2
n
1
xn    u n  1
2
49 2020/12/30
 3.3 The inverse z-Transform
3.3.2 Partial Fraction Expansion
M M M

b z k
k
z N
b z k
M k
b0  1  c z k
1

X  z  k 0
N
 k 0
N
 k 0
N

 1  d z 
a0
 k
a
k 0
z k
z M  ak z N  k
k 0 k 0
k
1

N
Ak
 1
if M  N
k 1 1  d k z

where Ak  1  d k z 1
 X  z  z d
50 2020/12/30 k
 Example 3.8
Second-Order z-Transform
1 1
X  z  , z 
 1 1  1 1  2
1  z 1  z 
 4  2 
A1 A2
X z   
 1 1   1 1 
1  z  1  z 
 4   2 
 1 1 
A1  1  z  X z   1
 4  z
1
 1 1 
4 A2  1  z  X z   2
 2  z
1
51 2020/12/30 2 Univ.
Zhongguo Liu_Biomedical Engineering_Shandong
 Example 3.8
Second-Order z-Transform

1 2 1
X z    , z
 1 1   1 1  2
1  z  1  z 
 4   2 

n n
1 1
xn  2  un    un
2 4

52 2020/12/30
 Example 3.9:
Inverse by Partial Fractions

X z  
1
1  2z  z

2

1 z 1 2
 , z 1
3 1 1 2  1 1 
1 z  z
2 2

1  z  1  z
1

 2 

53 2020/12/30
 1 2
1 2z  z A1 A2
X  z   B0  
3 1 1 2 1 1 1  z 1
1 z  z 1 z
2 2 2
2
1 2 3 1 2 1
z  z 1 z  2z 1
2 2 2 1
z  3z  2
1
5z 1
1
 1  5z
X z   2 
 1 1 

1  z  1  z
1

 2 
54 2020/12/30
 1  5 z 1 A1 A2
X  z  2   2 
 1 1  1  1

1  z  1  z 
1
1 z 1 1 z
 2  2
 
  1  5z 1
 
1 1 
A1    1  z   9

  2    
 1  1 z 1  1  z 1  2 
 1
z
2
 
  1  5z 1 
A2    1  z   8
 1

 1  1 z 1 1  z 1  

  2   
55 2020/12/30 z 1
 9 8
X  z  2   , z 1
 1 1  1  z 1 
1  z 
 2 
n
1 1
2  2 n
Z   un
Z

 1 1   2
1  z 
1  2 

Z
u n 
1 z 1
 n
1
xn  2 n  9  un  8un
2
56 2020/12/30
 LTI system Stability, Causality, and ROC
Review xn yn
hn
For a LTI system with impulse response h[n],
if it is causal, what do we know about h[n]?
Is h[n] one-sided or two-sided sequence?
Left-sided or right-sided?
Then what do we know about the ROC of the
system function H (z)?
If the poles of H (z) are all in the unit circle,
is the system stable?
57 2020/12/30
 LTI system Stability, Causality, and ROC
For H (z) with the poles as shown in figure ,
1
H  z 
Review
1  az 1  bz 1  cz 
1 1 1

can we uniquely determine h[n] ?

If ROC of H(z) is as shown Unit-circle
included
in figure, can we uniquely
determine h[n] ?

is the system stable ?

2020/12/30 Zhongguo Liu_Biomedical Engineering_Shandong Univ.

 LTI system Stability, Causality, and ROC
For H (z) with the poles as shown in figure ,
1
Review H  z 
1  az 1
1  bz 1
1  cz 1

If the system is causal
(h[n]=0,for n<0,right-sided ),
What’s the ROC like?

If ROC is as shown in

figure, is h[n] one-sided or
two-sided? Is the system
causal or stable?
2020/12/30 Zhongguo Liu_Biomedical Engineering_Shandong Univ.
 3.3 The Inverse Z-Transform
Review Inspection method
Partial fraction expansion
Power series expansion
 n  1, ROC : all z
1
a u n 
n
1
, ROC : z  a
1  az
1
 a u  n  1 
n
1
, ROC : z  a
1  az
60 2020/12/30
 Example 3.10:
Finite-Length Sequence
2  1 1 
  1

2 1 1 1
X  z   z 1  z  1  z 1  z  z  z  1  z
1

 2  2 2
 1, n  2
 1
 , n  1
 2
xn    1, n  0
 1
 2 , n 1
 0, otherwise

1 1
xn   n  2   n  1   n   n  1
61 2020/12/30 2 2
 Ex. 3.11: Inverse Transform by power
series expansion
X  z   log 1  az 1
, z a
n 1 n

(1) x
log 1  x   x  x  x  x 
1
2
2 1
3
3 1
4
4

n 1 n
x 1 n 1 n n

(1) a z
X  z   , az 1
1
n 1 n
 n 1 a
n
 
 1 , n  1 n 1 a
n
x ( n)   n   1 u  n  1
 0, n
 n 1
62 2020/12/30
 Example 3.12: Power Series
Expansion by Long Division
1
X z  
Z
1
, z  a  a u  n
n
1  az
1 2 2
1  az  a z
1  az 1
1 1 1 2 2
1
 1  az  a z  
1  az 1 1  az
1
az
1
az  a z 2 2 xn  a un
n

2 2
63 2020/12/30
a z 
 Example 3.13: Power Series Expansion for
a Left-sided Sequence
1
X z  
Z
1
, za  a u  n  1
n
1  az
a 1 z  a 2 z 2
1 1 2 2
az 1  1 1 1
  a z  a z 
1  az
1  a 1 z
1
a z
1
a za z 2 2 xn  a u n 1
n

a 2 z 2
64 2020/12/30
 Example 3.14:
Shifted Exponential Sequence
n
X z  
1 1  1
, z    u  n
Z
1 4 4
z
4  
1  1 
X z  
z
 4 
4 X z   z 
1

1 1 1 1  1  1 z 1 
1 z 1 z  
4 4  4 
n 1
1
xn    un  1
n
1
x  n   4  n   4   u  n 
4 4
65 2020/12/30
 Example 3.15:
Exponential Multiplication
1
u  n   Z
1
, z 1
1 z
x  n  r cos  w0 n  u  n 
n

1 1  jw0
  
u  n   re  u  n
n n
 re jw0

2 2
1
1
  u  n   2
n
re jw0 Z
jw0 1
, z r
2 1  re z
66 2020/12/30
 1 1  jw0
x  n   re   
u  n   re  u n
jw0 n n

2 2
1
1  jw0
  u  n  2
n
re Z
 jw0 1
, z r
2 1  re z
1 1
X  z  2  2
jw0 1  jw0 1
1  re z 1  re z


1  r cos w0 z 
1

, z  r
1 2 2
1  2r cos w0 z  r z
67 2020/12/30
 Example 3.17: Second-Order Pole
xn  na un  n a un
n
 n

1
a un
n
1
Z
, za
1  az
d  1 
X z    z  1 
, za
dz  1  az 
1
az
 , za
1  az  1 2

1
az
na un
n Z
, z a
68 2020/12/30
1  az  1 2