INVERSION OF THE Z-TRANSFORM

1. Direct Evaluation of
EEE 35: Signals & Systems
1
X (n ) = X ( z )z n 1dz
Lecture 4D 2j c
by contour integration over a closed path C that
Inverse Z-Transform encloses the origin and lies within the ROC.

C can be taken as a circle in the ROC of X(z) in

the z-plane.
4D-1 4D-2
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

2. By power series expansion

ROC
Expand X(z) into a power series of the form:

X (z ) = C z n
n

n=N

which converges in the given ROC.

C
By the uniqueness of Z-transform x(n) = Cn
for all n.
C is a contour that encloses the origin and lies
within the region of convergence of X(z). * If X(z) is rational, expansion is performed by
long division.
4D-3 4D-4
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

1
Example: Determine the inverse z-transform of Divide numerator by the denominator using long division
1 + 32 z 1 + 74 z 2 + 15
8
z 3 + ...
1
X (z ) =
1 1.5z 1 + 0.5z 2 1 23 z 1 + 12 z 2 1
1 23 z 1 + 12 z 2
if a) ROC : z > 1
b) ROC : z < 0.5
3
2
z 1 12 z 2
3
2
z 1 49 z 2 + 43 z 3
a) x(n) is a causal signal
Note:
7
4
z 2 43 z 3
seek power series expansion in negative X(z) should converge 7
4
z 2 21
8
z 3 + 78 z 4
powers of z at |z| > 1
15
8
z 3 78 z 4
4D-5 and so on 4D-6
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

3 1 7 2 15 3 31 4 b) x(n) is anticausal
X (z ) = 1 + z + z + z + z + ...
2 4 8 16 Power series expansion in positive powers of z
Divide this way:
3 7 15 31
x (n) = 1, , , , , ... 2 1
2z2 + 6z3 + 14z4 + 30z5 + ...
2 4 8 16 1
2
z z
3
2
+1 1
1 - 3z + 2z2
3z 2z2
x(n) may be written in closed form:
invert the order of
3z 9z2 + 6z3
the polynomials
2n +1 1 7z2 6z3
x (n) = u(n)
2n 7z2 21z3 + 14z4
15z3 14z4
Electrical and Electronics Engineering Institute EEE 35
4D-7
Electrical and Electronics Engineering Institute EEE 35 and so on 4D-8

2
 X(z) = 2z2 + 6z3 + 14z4 + 30z5 + ...
3. By partial fraction expansion
ROC : x < 0.5
Express X(z) as a linear combination
x (n) = {..., 30, 14, 6, 2, 0, 0}

X (z ) = 1 X1 (z ) + 2 X 2 (z ) + ... + K X K (z )
x(n) may be written in closed form:
if X1 (z ),..., X K (z ) x1 (n),..., x K (n)
1
z
n
x (n) = (2 2)u( n 1)
then x (n) = 1 x1 (n) + 2 x 2 (n) + ... + K x K (n)
* In general, a closed form solution is not possible.
Hence, this method is used only to determine the values
Write X(z) as a linear combination of Xk(z)s
of the first few samples of the signal. However, if the
pattern of the samples is simple enough, a closed form with known inverse transforms. (Refer to
x(n) may be inferred. table of transforms).
4D-9 4D-10
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

X(z) must be expressed as a rational function Example: Express the improper rational function
before partial fraction expansion can be performed 1 + 3z 1 + 11 z 2 + 13 z 3
X (z ) = 6
1 + 56 z 1 + 16 z 2
N(z ) b0 + b1z 1 + ... + bMz M
X (z ) = =
D(z ) 1 + a1z 1 + ... + aNz N in terms of a polynomial and a proper function.

* rational X(z) is called PROPER if aN 0 and M < N need to eliminate z-2 and z-3 in the numerator

* rational X(z) is called IMPROPER if M N do long division with the 2 polynomials

reversed & stop the division when the
can be written as sum of a polynomial order of the remainder becomes z-1
and a proper rational function
4D-11 4D-12
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

3
 2z 1 + 1 * In general any improper rational function
2 1 3 2
1
6
z + z
5
6
+1 1
3
z + 11
6
z + 3z 1 + 1 can be expressed as a sum of a polynomial
and a proper rational function
1
3
z 3 + 10
6
z 2 + 2z 1
1
6
z 2 + z 1
N(z) N (z)
z 2 + 56 z 1 + 1 X(z) = = C0 + C1z1 + ...+ CMNz(MN) + 1
1

D(z)
6
D(z)
remainder
1
6
z 1

Therefore: proper
1 z 1
X (z ) = 1 + 2z 1 + 6
5 1
1+ 6
z + 16 z2
4D-13 4D-14
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

In a proper rational function: Before performing partial fraction expansion, divide

N1 (z ) b0 + b1z + ... + bMz 1 M both sides by z (Borrow z from the right side. It
X (z ) = = will be returned later).
D(z ) 1 + a1z 1 + ... + aNz N
aN 0 & N > M
X (z ) b0 zN 1 + b1zN 2 + ... + bMzN M1
=
z zN + a1zN 1 + ... + aN
Eliminate the negative powers of z by multiplying
both numerator and denominator by zN In partial fraction expansion, express the proper
N N 1 NM rational function as a sum of simple fractions.
b0 z + b1z + ... + bMz
X (z ) = N N 1
z + a1z + ... + aN
* The denominator polynomial is factored in to
The result contains only positive powers of z. factors containing the poles p1, p2, pN of X(z)
4D-15 4D-16
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

4
 CASE 1: DISTINCT POLES
A1 z A z A z
X(z) = + 2 + ... + N
X (z) A1 A2 AN z p1 z p 2 z pN
= + + ... +
z z p1 z p2 z pN
A1 z A z A z
X(z) = + 2 + ... + N
(z pk ) X (z) z p1 z p2 z pN
Ak =
z z = pk
1 1 1
X(z) = A1 1
+ A2 1
+ ... + A N
Note: complex conjugate poles result in complex conjugate 1 p1z 1 p2z 1 pNz 1
coefficients in the PFE

After solving the coefficients A1 to AN, multiply both Inverse transformation will be performed to
sides by z (return the borrowed z to the right side of each term (use table of transforms).
the equation).
4D-17 4D-18
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

If there are complex-conjugate poles: CASE 2: MULTIPLE ORDER POLES

If A k = A k e jk pk = rk e jk If X(z) has a pole pk with multiplicity m, that is,

there is a (z - pk)m term at the denominator,
Ak
Z 1 1
+
A *k
* 1
[
= A k (pk )n + A *k (p*k )n u(n) ] then, the PFE should contain the terms
1 pk z 1 pk z
A1m A 2m A mk
+ + ... +
= 2 Ak rkn cos(k n + k )u(n) z pk (z pk )2
(z pk )m

A2m, A3m, , Amk are determined using

Note: C ejt = C(cos t + j sin t) differentiation.
4D-19 4D-20
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

5
 Example 1
Factor the denominator
Example 1: Determine the inverse z-transform of
X (z) z z
= 2 =
1 z z 1.5z + 0.5 (z 1)(z 0.5)
X(z) = ROC : | z | > 1
1 1.5z 1 + 0.5z 2
The partial fraction expansion is

Eliminate the negative powers of z by multiplying z2 to X(z) z A A2

= = 1 +
both numerator and denominator z (z 1)(z 0.5) z 1 z 0.5
z2
X (z) = 2
z 1.5z + 0.5 X (z) z(z 1)
A1 = (z 1) = =2
Borrow z from the right side of the equation
z z =1 (z 1)(z 0.5) z =1

X(z) z X (z) z(z 0.5)

= 2 A2 = (z 0.5) = = 1
z z 1.5z + 0.5 z z = 0.5 (z 1)(z 0.5) z = 0.5
4D-21 4D-22
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

Example 1 Example 1
X(z) 2 1
=
z z 1 z 0 .5 Since ROC of X(z) is |z| > 1, then
2 1
Return the borrowed z X(z) =
1 z 1 1 0.5z 1
2z z
X (z) =
z 1 z 0.5
Z-1
2 1
X(z) =
1 z 1 1 0.5z 1
Recall:
x (n) = [2(1) (0.5)n ]u(n) = [2 (0.5)n ]u(n)
n

Z 1
1
=
u(n) ( )
pnk
ROC: |z| > |pk| (causal)
1
( )
1 pk z - pnk u( n1) ROC: |z| < |pk|
(anticausal)
4D-23 4D-24
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

6
 Example 3
Ex. 3 Find the inverse z-transform of
Example 2: Determine the inverse z-transform of
z 3 + 3z 4
1 X (z) =
X(z) = ROC z|>
ROC:: |0.5 <10.75
|z| < 1 1 + 6z 1 + 14z 2 + 16z 3 + 8z 4
1 2
1 1.5z + 0.5 z

2 1 Multiplying z4 to both numerator and denominator:

X(z) = 1

1z 1 0.5z 1 z+3
X (z) =
z 4 + 6z3 + 14 z2 + 16z + 8
Z-1
A z cannot be borrowed, so divide both sides by z:
X(z) z+3
=
x(n) = -2(1)nu(-n-1) 0.5nu(n) z z(z + 6z + 14 z2 + 16z + 8)
4 3

4D-25 4D-26
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

Example 3 Example 3
X (z) A1 A A3 A4 A *4
= + 2 + + +
The roots of the polynomial enclosed by the parenthesis are: z z z + 2 (z + 2)2 z + 1 j z + 1 + j
-2, -2, -1 + j, -1 - j

zX(z) z(z + 3) 3
The PFE would be A1 = = =
z z = 0 z(z + 2)2 (z + 1 j)(z + 1 + j) z = 0 8
X(z) z+3
=
z z(z + 6z + 14z2 + 16z + 8)
4 3
(z + 2)2 X(z) (z + 2)2 (z + 3) 1
A3 = = 2
=
z+3
z z = 2
z( z + 2) ( z + 1 j)( z + 1 + j) z = 2
4
=
z(z + 2)2 (z + 1 j)(z + 1 + j)

A A A3 A4 A *4 d (z + 2)2 X(z) d (z + 3) 5
= 1 + 2 + + + A2 = ( ) =( ) =
z z + 2 (z + 2)2 z + 1 j z + 1 + j dz z dz z(z + 1 j)( z + 1 + j) z = 2
8
z = 2

4D-27 4D-28
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

7
 Example 3 Example 3
Consequently, 3 58 1 4 z 1 1 + j3 1 j3
X (z) = + 1
+ 1 2
+ 8 8
1
+ 8 8
8 1 + 2z (1 + 2z ) 1 + (1 j)z 1 + (1 + j)z 1
A4 = 1
8
+ j 38 A *4 = 1
8
j 38
Therefore 3 58 1 8 (2z 1 ) 1 + j3 1 j3
X (z) = + 1
+ 1 2
+ 8 8
1
+ 8 8
8 1 + 2z (1 + 2z ) 1 + (1 j)z 1 + (1 + j)z 1
X (z) 3 8 5 8 14 1 + j3 1 j3
= + + + 8 8
+ 8 8
z z z + 2 (z + 2) 2
z +1 j z +1 + j The first term is The second term is
x1 (n) = (n) 3 x 2 (n) = 58 (2)n u(n)
3 58 z 14 z ( 1 + j 3 8 )z ( 1 8 j 3 8 )z 8
X (z) = + + + 8 +
8 z + 2 (z + 2) 2
z +1 j z +1 + j
The third term is From the table, recall that
3 58 1 4 z 1 1 + j3 1 j3
X (z) = + + + 8 8
+ 8 8 x3 ( n) = n(2) u (n)
1 n
z az 1
8 1 + 2z 1
(1 + 2z )1 2
1 + (1 j)z 1
1 + (1 + j)z 1
8 n anu(n)
(1 az 1 )2
4D-29 4D-30
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

Example 3
For the last two terms, recall that Recall
Fourier Series for DT Periodic Signals
Ak
Z 1
1 p z 1
+
1
A *k
p * 1
z
[
= A k (pk )n + A *k (p*k )n u(n) ] N 1
k k
= 2 A k rkn cos(k n + k )u(n)
x(n) = k = 0
cke j2 n / N
, N = period

Since 1 N -1

Ak = 1
8
+ j 38 = 0.3953e j1.249 and p k = 1 + j = 1.414 e j2.3562 ck =
N

n=0
x(n) e - j2 kn/N

then
x 4 (n) = 2(0.3953)1.414n cos(2.3652n + 1.249) ! The spectrum of a signal x(n), which is periodic with
period N, is a periodic sequence with period N.
= 0.7906(1.414)n cos(2.3652n + 1.249)
4D-31 4D-32
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

8
Recall
Fourier Transform for DT Aperiodic Signals Convolution Theorem
- j n
X( ) = x(n) e Y() = H() X() =F { h(n) * x(n)}
n =

1 j n
x(n) = X( )e d
2 2 LTI
X() H() Y()
Periodicity of X():
X()=X( + 2k)
4D-33 4D-34
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

Recall Z-Transform Definition Z Transform & Fourier Transform

X(z) = [x(n)z
x(n)r -n ]
n =-
-n
If X(z) converges for |z| = 1, then

X(z) z=e
j = X() = x(n) e-jn
n =-
Let z = rej w

X(z) = [x(n)r ] e-jn

n =-
-n

FT = ZT unit circle if |z| = 1 ROC

= FT {x(n) r-n}
4D-35 4D-36
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

9
 Z Transform & Fourier Transform

1 Example
Ex. ZT{an u(n)} = , |a| > 1
1 - az -1
does not have a FT Let the sequences y(n) and x(n) be related through

y(n) - 0.25 y(n-2) = x(n)

Ex. x(n) = sin C n , a finite energy signal, a. Find the frequency response for a system G() such that
n
has no ZT, g*y = x

1 < C
but X() = b. What is the impulse response of the system G?
0 C <
4D-37 4D-38
Electrical and Electronics Engineering Institute EEE 35 Electrical and Electronics Engineering Institute EEE 35

Example

Find the magnitude and phase responses of

the system described the LCCDE

y(n) = x(n) + 0.5 x(n-p) + 0.5 x(n+p)

where x(n) is the input sequence.

4D-39
Electrical and Electronics Engineering Institute EEE 35

10