Inverse Z-Transform

Dr Malaya Kumar Hota (Prof., SENSE, VIT University)

Inverse z-transform
In general, the inverse z-transform may be found by using any
of the following methods:
 Direct evaluation by contour integration (or residue method)
1

n 1
x ( n)  X ( z ) z dz
2j C

 Power series expansion method (or Long division)

 Partial fraction expansion method.

DR. MALAYA KUMAR HOTA (PROF., SENSE) 1

Power series expansion method (or
Long division)
z 2  2 z 1  2
Determine the inverse z-transform of X ( z)  1
; z  1
z 1
Perform long division:
2  z 2  z 3  z 4  ...
1 z 1| 2  2z 1  z 2
2  2 z 1
z 2
z 2  z 3

X ( z )  2  z 2  z 3  z 4  ......

x(n)  {2,0,1,
  1,1,...}

DR. MALAYA KUMAR HOTA (PROF., SENSE) 2
 Power series expansion method (or
Long division)
1
Determine the inverse z-transform of X ( z) 
1  1.5z 1  0.5z 2

when (a) ROC: z  1 (b) ROC: z  0.5

(a) Since ROC is exterior of a circle, we expect x(n) to be a causal signal.

Thus we seek a power series expansion in negative power of z.

(b) Since ROC is interior of a circle, we expect x(n) to be a anti-causal signal.

Thus we seek a power series expansion in positive power of z.

DR. MALAYA KUMAR HOTA (PROF., SENSE) 3

 Power series expansion method (or
Long division)
1 32 z 1  74 z 2  158 z 3  ...
1 32 z 1  12 z 2| 1
1 3 z 1  1 z 2
2 2 3 7 15
3 1 1 2 X ( z )  1  z 1  z 2  z 3  ......
z  z 2 4 8
2 2
3 z 1  9 z 2  3 z 3
3 7 15
2 4 4 x(n)  {1, , , ,...}
7 2 3 3 2 4 8
z  z
4 4
7 z 2  21 z 3  7 z 4
4 7 8
15 3 7 4
8
z 8 z 

DR. MALAYA KUMAR HOTA (PROF., SENSE) 4

Power series expansion method (or
Long division)
2z2  6z3 14z4  30z5 ...
1 2 3 1
2
z  2 z 1 | 1
1  3z  2 z 2
3z  2 z 2
3z  9 z 2  6 z 3
7 z 2  6 z3
7 z 2  21z3 14 z 4
15z3 14z4

X ( z )  2 z 2  6 z 3  14 z 4  30 z 5  ......

 x(n)  {...,30,14, 6, 2, 0, 0}



DR. MALAYA KUMAR HOTA (PROF., SENSE) 5

Power series expansion method (or
Long division)
1. The method of long division will not provide answers
for x(n) when n is large because the long division
becomes tedious.
2. Although, the method provides a direct evaluation of
x(n), a closed form solution is not possible, except if the
resulting pattern is simple.
3. Hence this method is used only if one wished to
determine the values of first few samples of the signal.

DR. MALAYA KUMAR HOTA (PROF., SENSE) 6

 Partial fraction expansion method
1
Example: Determine the inverse z-transform of X ( z ) 
1  1.5z 1  0.5z 2
when (a) ROC: z  1 (b) ROC: z  0.5 (c) ROC:  z  1

Solution:
1
X ( z) 
1  1.5z 1  0.5z 2
To eliminate the negative power, multiply numerator and denominator by z2.

z2
X ( z)  2 Not a proper function
z  1.5z  0.5

DR. MALAYA KUMAR HOTA (PROF., SENSE) 7

Partial fraction expansion method
X ( z) z
 2 Proper function
z z  1.5 z  0.5
z

( z  1)( z  0.5)
A B
 
( z  1) ( z  0.5)

z 1
A  2
( z  0.5) z  1 1  0.5
z 0.5
B   1
( z  1) z  0.5 0.5  1

DR. MALAYA KUMAR HOTA (PROF., SENSE) 8

Partial fraction expansion method
X ( z) 2 1
 
z ( z  1) ( z  0.5)

2z z
X ( z)   Poles at p1  1and p2  0.5
( z  1) ( z  0.5)
2 1
 
(1  z 1 ) (1  0.5 z 1 )

(a) ROC: z  1 The signal x(n) is causal.

 x(n)  2 1 u  n    0.5  u  n 

n n

  2   0.5   u  n 
n
 

DR. MALAYA KUMAR HOTA (PROF., SENSE) 9

Partial fraction expansion method
(b) ROC: z  0.5 The signal x(n) is anti-causal.

 x(n)  2 1 u  n  1   0.5  u  n  1

n n

 2   0.5  u  n  1
 n
 

(c) ROC:  z  1 The signal x(n) is non-causal.

 x(n)  2 1 u  n  1   0.5 u  n 

n n

DR. MALAYA KUMAR HOTA (PROF., SENSE) 10

Causality and Stability
 A causal LTI system is one whose unit sample
response, h(n)=0 for n<0.
 A LTI system is causal if and only if the ROC of the
system function is the exterior of the circle of radius
r<.
 A LTI system is BIBO stable if and only if the ROC of
the system function includes the unit circle.

DR. MALAYA KUMAR HOTA (PROF., SENSE) 11

 Causality and Stability
Example: A linear time-invariant system is characterized by the system function

3  4 z 1
H ( z) 
1  3.5 z 1  1.5 z 2

Specify the ROC of H(z) and determine h(n) for the following conditions:
(a) The system is stable.
(b) The system is causal.
(c)The system is anti-causal.

DR. MALAYA KUMAR HOTA (PROF., SENSE) 12

 Causality and Stability
Solution: 3  4 z 1
H ( z) 
1  3.5 z 1  1.5 z 2
To eliminate the negative power, multiply numerator and denominator by z2.

3z 2  4 z
H ( z)  2 | Not a proper function
z  3.5 z  1.5
H ( z) 3z  4
 2 Proper function
z z  3.5 z  1.5 3z  4 94
A  2
3z  4 ( z  0.5) z  3 3  0.5

( z  3)( z  0.5)
A B 3z  4 1.5  4
  B  1
( z  3) ( z  0.5) ( z  3) z  0.5 0.5  3

DR. MALAYA KUMAR HOTA (PROF., SENSE) 13

 Causality and Stability
H ( z) 2 1
 
z ( z  3) ( z  0.5)
2z z
H ( z)   Poles at p1  3and p2  0.5
( z  3) ( z  0.5)
2 1
 
(1  3 z 1 ) (1  0.5 z 1 )

(a) Since the system is stable, its ROC must includes the unit circle and hence it is
  z  3 . Consequently, h(n) is non-causal and is given as

h(n)   0.5 u  n   2  3 u  n  1
n n

DR. MALAYA KUMAR HOTA (PROF., SENSE) 14

 Causality and Stability
(b) Since the system is causal, its ROC is  z  3

h(n)   0.5  u  n   2  3 u  n 
n n

This system is unstable.

  0.5   2  3  u  n 
n n
 

(c) Since the system is anti-causal, its ROC is  z  0.5

h(n)    0.5  u  n  1  2  3 u  n  1
n n

   0.5   2  3  u   n  1
n n This system is unstable.
 

DR. MALAYA KUMAR HOTA (PROF., SENSE) 15

Pole-Zero Cancellations
Example: Consider the system:

1  2 z 1  2 z 2  z 3
H ( z)  ROC: 0.5< z  1
1  z 1 1  0.8z 1 1  0.4 z 1 
(a) Sketch the pole-zero pattern. Is the system stable.
(b) Determine the impulse response of the system.

DR. MALAYA KUMAR HOTA (PROF., SENSE) 16

Pole-Zero Cancellations
Solution:

1  2 z 1  2 z 2  z 3
H ( z) 
1  z 1 1  0.8z 1 1  0.4 z 1 

1  z 1  z  z 
1 1 2

Pole zero cancellations

1  z 1  0.8z 1  0.4 z 
1 1 1


1 z  z 
1 2

1  0.8z 1  0.4 z 

1 1


 z  z  1
2

 z  0.8 z  0.4 

DR. MALAYA KUMAR HOTA (PROF., SENSE) 17

Pole-Zero Cancellations
(a) Poles: p1=0.8 and p2=0.4.
Zeros:
1 3
z1,2  j
2 2
Causal system is stable since poles are within unit circle.

(b)
H ( z) 
z 2
 z  1
 z  0.8 z  0.4 

z 2
 z  1
z 2
 1.2 z  0.32  H(z) is not proper.

DR. MALAYA KUMAR HOTA (PROF., SENSE) 18

Pole-Zero Cancellations
Divide H(z) once to make proper function.

H ( z)  1 
 0.2 z  0.68
z 2
 1.2 z  0.32 

 1
 0.2 z  0.68
 z  0.8 z  0.4 
A B
 1 
 z  0.8  z  0.4 
0.2 z  0.68 0.16  0.68
A   2.1
( z  0.4) z  0.8 0.8  0.4
0.2 z  0.68 0.8  0.68
B   1.9
( z  0.8) z  0.4 0.4  0.8

DR. MALAYA KUMAR HOTA (PROF., SENSE) 19

Pole-Zero Cancellations
2.1 1.9
H ( z)  1  
 z  0.8  z  0.4 
2.1z 1 1.9 z 1
 1 
1  0.8z  1  0.4 z 1 
1

h(n)   (n)  2.1 0.8  u  n  1  1.9  0.4  u  n  1

n 1 n 1

  (n)   2.1 0.8   1.9  0.4   u  n  1

n 1 n 1
 

DR. MALAYA KUMAR HOTA (PROF., SENSE) 20