Topic13 Inversion of Z Transform
Topic13 Inversion of Z Transform
Replace z = rejΩ
∞
X
X(z) = x[n]r−n e−jΩn , z ∈ C (2)
n=−∞
−n
which is basically the DTFT of x[n]r . If r = 1 (i.e. unit circle), z -transform equals DTFT.
So we can say that DTFT of a sequence is equal to its z-transform evaluated along the unit
circle (if the ROC includes unit circle).
1
Solution: The answer to this at a = b would be
−5z/4
aX(z) + aY (z) = a .
(z − 1/4)(z − 3/2)
The ROC would be 1/4 < |z| < 3/2. The zero at 1/2 cancels the pole at 1/2. Since pole define
the ROC boundary, hence cancellation of pole results in larger ROC than the X(z).
3 Inversion of z -transform
The inverse of z− transform is often done using method of partial fraction. It is as simple as
we have seen for the Laplace transform case. first, we break a given rational polynomial into
partial fractions. And then, we write the time domain form with the help of known z transform
pairs. Based on the ROC, we can figure out whether our time domain sequence would be left
sided, right sided, or two sided.
Example 3: Find the inverse z transform of
1 − z −1 + z −2
X(z) =
(1 − 12 z −1 )(1 − 2z −1 )(1 − z −1 )
with ROC 1 < r < 2.
Solution: We can write
A B C
X(z) = 1 −1 + −1
+
1 − 2z 1 − 2z 1 − z −1
which leads to
1 2 2
X(z) = 1 −1 + −1
+
1− 2
z 1 − 2z 1 − z −1
Based on the given ROC, we can write:
1
n
x[n] = u[n] − 2(2)n u[−n − 1] − 2u[n]
2
• r < 1/2
Answers: 1 n
x[n] = u[n] − 2(2)n u[−n − 1] + 2u[−n − 1]
2
and, 1 n
x[n] = − u[−n − 1] − 2(2)n u[−n − 1] + 2u[−n − 1]
2
2
with ROC r < 1.
Solution: Step 1: Convert it in a polynomial ratio in z −1
1 1 − 10z −1 − 4z −2 + 4z −3
X(z) = z
2 1 − z −1 − 2z −2
Step 2: After division and partial fractions we can write
" #
1 1 1 3
X(z) = zW (z) = z − 2z −1 + 3 + −
2 2 1 + z −1 1 − 2z −1
we can directly compare the expression and figure out time domain representation. Let’s see
with examples:
Example 6: Find inverse z transform of
That is
x[n] = 4δ[n + 2] + 2δ[n] + 3δ[n − 1]
3
Thus after comparison from Eq. 1, we can write
( n+1
(−1) an
n
n≥1
x[n] =
0 n≤0
or equivalently
(−a)n
x[n] = − u[n − 1]
n
5 Transfer function
Consider a LTI system with input x, output y , and impulse response h, and let X, Y , and H
denote the z transforms of x, y, and h, respectively. Then,
Y (z) = H(z)X(z)
We refer to H as the system function (or transfer function) of the system.
Example 8: Find the impulse response h[n] of an LTI system if the input is
x[n] = (−1/3)n u[n]
and the output is
y[n] = 3(−1)n u[n] + (1/3)n u[n]
Solution: We can write z-transforms for input and output as
1
X(z) = ; r > 1/3
1 + (1/3)z −1
3 1
Y (z) = −1
+ ;r > 1
1+z 1 − (1/3)z −1
Now with the help of Y (z) = H(z)X(z) we can write
2 2
H(z) = −1
+ ;r > 1
1+z 1 − (1/3)z −1
h[n] = 2(−1)n u[n] + 2(1/3)n u[n]
4
6 Causality and stability
If an LTI system is causal, its impulse response is causal, and therefore right sided. i.e.
∞
X
H(z) = h[n]z −n
0
It does not include positive powers of z. From this, we have the result below:
1. In general, an LTI system is causal if the ROC of the system function H(z) is an outward
circle, including ∞.
2. For a rational H(z), the system is causal if and only if the ROC is exterior of the circle,
and, the order of numerator cannot be greater than the order of denominator.
An LTI system is BIBO stable if and only if the ROC of its system function H(z) contains
the unit circle. Try h[n] = 2−n u[n] or h[n] = −2n u[−n − 1].
In terms of poles:
• A causal system should have a region of convergence outside the outermost pole. A stable
system should have the unit circle in its region of convergence. Therefore, a system that
is both stable and causal must have all its poles inside the unit circle in the
z-plane.
Example 9: Is the following system causal and stable?
1 1
H(z) = −1
+ ;r > 2 (3)
1 − 0.5Z 1 − 2z −1
Answer: Since it’s in a rational form, check 2nd condition of causality. Yes it is causal but
not stable.
Example 10: Consider the signal x[n] = an u[n] and find unilateral z-transform.
Solution: Since x[n] = 0, for n < 0, the unilateral and bilateral transforms are equal :
1
X(z) = ; r > |a|
1 − az −1
Example 11: Consider the signal x[n] = an+1 u[n + 1] and find unilateral z-transform.
Solution: In this case the unilateral and bilateral transforms are not equal. The bilateral
transform can be written as (time shift property in above example):
z
X(z) = ; r > |a|
1 − az −1
In contrast, the unilateral transform is:
∞
X
X(z) = an+1 z −n
n=0
a
= ; r > |a|
1 − az −1
5
Reference material
1. Textbook: Signals and Systems by Simon Haykin
2. Lecture notes are inspired from the course materials of JHU 520.214; MIT 6.003; Purdue
ECE-301; UVic ECE-260; and Imperial College E2.5
[Please report any typos in the notes by sending an email to the instructor.]