Chapter 3

The z-Transform
Der-Feng Tseng
Department of Electrical Engineering
National Taiwan University of Science and Technology
(through the courtesy of Prof. Peng-Hua Wang of National Taipei University)

February 19, 2015

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

1 / 25

Outline

3.1 z-Transform

3.2 Properties of the ROC

3.3 The Inverse z-transform

3.4 z-transform Properties

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

2 / 25

3.1 z-Transform

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

3 / 25

Definition
Definition (z-transform) The z-transform of x[n] is defined as

Z{x[n]} = X(z) ,

x[n]z n

n=

The z-transform evaluated on the unit circle corresponds to the

Fourier transform. Let z = rej , we have
X(rej ) =

(x[n]r n )ejn .

n=

Fourier transform of x[n]r n .

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

4 / 25

ROC
Now we consider the region in which the z-transform is converge. It is
called region of convergence (ROC).
z-transform converges:
|X(z)| = |X(rej )|

|x[n]r n | < .

n=

Fourier transform converges:

|X(ej )|

|x[n]| < .

n=

It is possible that X(z) converges but X(ej ) does not. For example,
the z transform of the unit step function converges for |z| > 1.

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

5 / 25

Examples 3.1 3.6

Example 3.1 x[n] = an u[n], X(z) =

1
1az 1 ,

Example 3.2 x[n] = an u[n 1], X(z) =

|z| > |a|.

1
,
1az 1

|z| < |a|.

Example 3.3 x[n] = ( 12 )n u[n] + ( 31 )n u[n],

X(z) =

1
)
2z(z 12
,
1
(z 2 )(z + 31 )

1
|z| > .
2

Example 3.5 x[n] = ( 31 )n u[n] ( 12 )n u[n 1]

Example 3.6 x[n] = an (u[n] u[n N ])

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

6 / 25

3.2 Properties of the ROC

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

7 / 25

Summary of Properties
1

The ROC is a ring or disk centered at the origin;

0 rR < |z| < rL

The Fourier transform of x[n] exists if and only if the ROC includes
the unit circle.

The ROC cannot contain any poles.

If x[n] is a finite-duration sequence, then the ROC is the entire

z-plane; except possible z = 0 or z = .

If x[n] = 0, n < N1 < is a right-side sequence, then the ROC

extends outward from the outermost finite pole to z = .

If x[n] = 0, N2 < n is a left-side sequence, then the ROC extends

inward from the innermost nonzero pole to z = 0.

A two-side sequence is an infinite-duration sequence. The ROC of a

two-side sequence will consist of a ring in the z-plane.

The ROC must be a connected region.

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

8 / 25

Properties 1 & 2
Property 1. The ROC is a ring or disk centered at the origin.
The convergence of
X(z) ,

x[n]z n

n=

depends on only |z|.

Property 2. The Fourier transform of x[n] exists if and only if the ROC
includes the unit circle.
X(z) reduces to the Fourier transform when |z| = 1.
System is stable. h[n] is absolutely summable. Fourier
transform exists. ROC includes the unit circle.

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

9 / 25

Properties 3 & 4
Property 3. The ROC cannot contain any poles.
X(z) does not converge at poles.
Property 4. If x[n] is a finite-duration sequence, then the ROC is the
entire z-plane; except possible z = 0 or z = .
If r 6= 0, ,
j

|X(z)| = |X(re )|

M2
X

|x[n]r n | < .

n=M1

For example, let X1 (z) = 1 + z 1 , X2 (z) = 1 + z, and

X3 (z) = z + 1 + z 1 . We have ROC1 = Z {0}, ROC2 = Z {},
ROC3 = Z {0, }.

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

10 / 25

Property 5
Property 5. If x[n] = 0 for n < N1 is a right-side sequence, then the
ROC extends outward from the outermost finite pole to z = .
P
n
Suppose x[n] = N
k=1 Ak dk , we need

n=N1

or

N

X

X


Ak dnk r n  <
|x[n]r n | =



n=N1 k=1

|dnk r n | <

n=N1

for k = 1, 2, . . . , N . Thus, |dk r 1 | < 1, i.e., r > |dk |

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

11 / 25

Property 6
Property 6. If x[n] = 0, N2 < n is a left-side sequence, then the ROC
extends inward from the innermost nonzero pole to z = 0.
P
n
Suppose x[n] = N
k=1 Ak dk , we need
N2
X

n=

or

N

N2  X

X


Ak dnk r n  <
|x[n]r n | =



n= k=1

N2
X

|dnk r n | <

n=

for k = 1, 2, . . . , N . Thus, |dk r 1 | > 1, i.e., r < |dk |

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

12 / 25

Example 3.7
If a system has the zero-pole plot shown as follows. There are three
possible ROCs.

System 1: |z| > 2: not stable!, causal!

System 2: |z| < 1/2: not stable! not causal!
System 3: 1/2 < |z| < 2: stable, not causal!
Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

13 / 25

3.3 The Inverse z-transform

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

14 / 25

Method 1: Inspection Method

Find the sequence x[n] with the following z-transform
X(z) =

1
1

1 1 ,
2z

1
|z| > .
2

Solution:

X
1
|z| > X(z) =
2

n=0

Der-Feng Tseng (NTUST)

 n
 n
1
1
n
z x[n] =
u[n]
2
2

DSP Chapter 3

February 19, 2015

15 / 25

Method 2: Partial Fraction Expansion

Let

X(z) =

M
X

bk z k

k=0

N
X

ak z k

k=0

Case 1. M < N . All poles are simple.

X(z) =

N
X
k=1

Der-Feng Tseng (NTUST)

Ak
,
1 dk z 1


Ak = (1 dk z 1 )X(z)z=d

DSP Chapter 3

February 19, 2015

16 / 25

Method 2: Partial Fraction Expansion

Case 2. M N . All poles are simple.
X(z) =

M
N
X

Br z r +

r=0

N
X
k=1

Ak
1 dk z 1

Case 3. M N . X(z) has a s-order pole di

X(z) =

M
N
X
r=0

Br z

N
X
k=1
k6=i

s
X
Ak
Cm
+
1
1 dk z
(1 di z 1 )m
m=1

[n r] z r

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

17 / 25

Example 3.8
Find the sequence x[n] with the following z-transform
X(z) =

1
(1

Solution:
X(z) =

1 1
4 z )(1

1 1 ,
2z )

|z| >

1
2

A1
A2
1 1 +
1 4z
1 12 z 1

where A1 = 1, A2 = 2.
x[n] = (2 2n 4n )u[n].

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

18 / 25

Example 3.9
Find the sequence x[n] with the following z-transform
X(z) =

(1 + z 1 )2
,
(1 12 z 1 )(1 z 1 )

Solution:
X(z) = B0 +

|z| > 1

A2
A1
+
1 21 z 1 1 z 1

where B0 = 2, A1 = 9, A2 = 8.
x[n] = 2[n] (9 2n + 8)u[n].

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

19 / 25

Method 3: Power Series Expansion

X(z) = + x[2]z 2 + x[1]z + x[0] + x[1]z 1 + x[2]z 2 +

x[n] = + x[2][n + 2] + x[1][n + 1] + x[0][n]
+ x[1][n 1] + x[2][n 2] +
Example 3.10
1
X(z) = z 2 (1 z 1 )(1 + z 1 )(1 z 1 )
2
1
1
= z 2 z 1 + z 1
2
2
1
1
x[n] = [n + 2] [n + 1] [n] + [n 1]
2
2

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

20 / 25

Method 3: Power Series Expansion

Example 3.11
X(z) = log(1 + az 1 ) =

X
(1)n+1 an z n

n=1

x[n] =

Der-Feng Tseng (NTUST)

(1)n+1 z n

0,

, n1
n 0.

DSP Chapter 3

February 19, 2015

21 / 25

3.4 z-transform Properties

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

22 / 25

Properties
Linearity.
Z{ax1 [n] + bx2 [n]} = aZ{x1 [n]} + bZ{x2 [n]} ROC = Rx1 Rx2
Time-Shifting.
Z{x[n n0 ]} = z n0 Z{x[n]}

ROC = Rx {z = or 0}

Multiplication. If Z{x[n]} = X(z), then

Z{z0n x[n]} = X(z/z0 )
with ROC = |z0 |Rx .
Differentiation. If Z{x[n]} = X(z), then
Z{nx[n]} = z

d
X(z)
dz

with ROC = Rx .
Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

23 / 25

Properties
Complex Conjugation. If Z{x[n]} = X(z), then
Z{x [n]} = X (z )
with ROC = Rx .
Time Reversal. If Z{x[n]} = X(z), then
Z{x [n]} = X (1/z )
with ROC = 1/Rx .
Time Reversal. If x[n] is real, then
Z{x[n]} = X(1/z)
with ROC = 1/Rx .

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

24 / 25

Properties
Convolution.
Z{x1 [n] x2 [n]} = Z{x1 [n]}Z{x2 [n]} ROC = Rx1 Rx2
Initial-Value Theorem. If x[n] is causal, then
lim X(z) = x[0]

Der-Feng Tseng (NTUST)

DSP Chapter 3

February 19, 2015

25 / 25