EEE-324 Digital Signal Processing

Lab#6: Z-Transform and Inverse Z-Transform

Name Roman Bin Ehsan

Reg no# FA22-BEE-258
Class BEE-6A
Instructor Dr. Bakhtiar Ali
Introduction to Z-Transform and Inverse Z-Transform
Overview

The Z-transform is an essential mathematical tool used in signal processing and

control systems to analyze discrete-time signals and systems. It serves as the discrete
counterpart of the Laplace transform, allowing for efficient representation and
manipulation of discrete-time signals in the complex frequency domain. This lab
session aims to introduce students to the computation of the Z-transform, its
properties, and its inverse, providing a comprehensive understanding of its
applications in analyzing Linear Time-Invariant (LTI) systems.

Objectives

One of the primary objectives of this lab is to familiarize students with the computation
of the Z-transform for various discrete signals and the determination of their Region of
Convergence (ROC). Understanding the ROC is crucial, as it dictates the stability and
causality of the system being analyzed. Students will also learn how the Z-transform
simplifies mathematical computations, making it a powerful tool for solving difference
equations that describe LTI systems.

Inverse Z-Transform

Additionally, this lab emphasizes the process of inverting the Z-transform using the
method of partial fraction expansion. The ability to determine the inverse Z-transform is
essential in reconstructing original signals from their transformed counterparts.
Through this approach, students will explore how LTI systems are represented in the Z-
domain and their relationship with frequency response. The lab will also highlight the
distinctions between bilateral and unilateral Z-transforms, demonstrating their
respective use cases in signal analysis.

Pole-Zero Plot and Stability Analysis

A critical aspect covered in this lab is the pole-zero plot, a graphical representation of
the poles and zeros of a system on the Z-plane. This visualization is instrumental in
understanding the stability and frequency response characteristics of a system. By
analyzing the pole-zero distribution, students will gain insight into the behavior of
discrete-time systems and their implications in real-world applications.

MATLAB Implementation

Furthermore, students will be introduced to MATLAB as a computational tool for

solving Z-transform problems efficiently. The software's built-in functions enable quick
and accurate calculations, particularly in determining the inverse Z-transform through
partial fraction expansion. This hands-on experience with MATLAB will reinforce
theoretical concepts and enhance problem-solving skills in digital signal processing.
In Lab Task
Task#1: Using MATLAB we determine the partial-fraction expansion of the z-
transform X(z) given by

MATLAB Code:

Numerator=[18, 0, 0, 0]; disp(r);

Denominator=[18, 3, -4, -1,]; disp('poles');
[r,p,k] = disp(p);
residuez(Numerator,Denominator); disp('Direct Term');
disp('Residue'); disp(k);

Output:

Discussion:

In Task 1, a rational function is decomposed into partial fractions using the residuez
function, a valuable tool in signal processing. The Numerator and Denominator
vectors store the coefficients of the polynomial expressions representing the function
in the z-domain. By applying residuez(N, D), the function determines the residues (r),
poles (p), and any constant term (k). The residues and poles contribute to the
function’s decomposition, with poles marking points of prominence in the z-plane.
Analyzing these values offers a deeper understanding of the function’s frequency
characteristics, aiding in digital filter design and signal processing applications.

Task#2: Determine the inverse z-transform of the 11 terms of following equation

MATLAB Code:

N=[1,-2]; disp('poles');
D=[1, 0.4, -0.12]; disp(p);
[r,p,k] = residuez(N,D); disp('Direct Term');
n_terms=11; disp(k);
impz(N,D,n_terms) disp('Terms');
disp('Residue'); disp(terms);
disp(r);
Output:

Graph:

Figure 1: Impulse response of Inverse Z-transform of 11 terms

Discussion:

In this task, a discrete-time system was examined by breaking down its transfer
function into partial fraction components using the residuez function. The transfer
function is characterized by the numerator coefficients N = [1, -2] and the
denominator coefficients D = [1, 0.4, -0.12]. By applying residuez(N, D), the system’s
 residues (r), poles (p), and direct term (k) were determined. The residues indicate the
contribution of each component at its corresponding pole, while the poles identify
locations in the z-plane where the system's response is most pronounced. The direct
term (k) accounts for any constant component in the system’s output.

Once the residues and poles were obtained, the system’s impulse response was
computed for the first 11 terms. This was accomplished using the filter function,
which processes an impulse signal—[1, zeros(1, n_terms-1)]—through the transfer
function defined by N and D. The resulting sequence, stored in terms, represents the
system’s behavior in the time domain during these initial samples. Finally, the
extracted residues, poles, direct term, and impulse response values were displayed,
offering valuable insights into the system’s stability, frequency response, and
transient behavior.

Task#3: Determine the Inverse z-transform of the following sequences, using

partial fraction expansion method.

MATLAB Code:

N=[1, -1, -4, 4]; disp(r);

D=[1, -2.75, 1.625,-0.25]; disp('poles');
[r,p,k] = residuez(N,D); disp(p);
impz(N,D) disp('Direct Term');
disp('Residue'); disp(k);

Output:
 Figure 2:Impulse response of the partial fraction for n terms

Discussion:

This code analyzes a discrete-time transfer function by performing a partial fraction

expansion and computing its impulse response. The system is defined by the
numerator coefficients N = [1, -1, -4, 4] and denominator coefficients D = [1, -2.75,
1.625, -0.25]. Using residuez(N, D), the code determines the system's residues (r),
poles (p), and direct term (k), breaking down the transfer function into simpler
components that help reveal its frequency response and stability.
To analyze the system’s time-domain behavior, the impz(N, D) function is used to
compute and visualize the impulse response, showing how the system reacts to an
impulse input over time. Finally, the extracted residues, poles, and direct term are
displayed, providing valuable insights into the system’s characteristics.

Task#4: A digital filter is described by the difference equation,

Using the freqz function, plot the magnitude and phase of the frequency response
of the filter.

MATLAB Code:

b = [1,1]; title('Magnitude Response');

a= [1, -0.9, 0.81]; xlabel('frequency in pi units');
w = [0:1:500]*pi/500; ylabel('?H?');
H=freqz(b,a,w); subplot(2,1,2);
magH = abs(H); plot(w/pi,phaH);
phaH = angle(H)*180/pi; title('Phase Response');
subplot(2,1,1); xlabel('frequency in pi units');
plot(w/pi,magH); ylabel('Degrees');
Output:

Figure 3:Magnitude and Phase of the frequency response of the filter.

Discussion:

This task involves analyzing the frequency response of a digital filter described by the
given difference equation. The filter is implemented using the coefficients b = [1,1] for
the numerator and a = [1, -0.9, 0.81] for the denominator. The freqz function
computes the frequency response over 501 points between 0 and π radians/sample.
The magnitude response, obtained from the absolute values of H, shows how the filter
affects different frequency components, while the phase response, derived from the
angle of H, reveals how the filter alters the phase of incoming signals. The results are
plotted, with the magnitude response displayed in the first subplot and the phase
response in the second, providing insight into the filter’s frequency-dependent
behavior.

From the magnitude response, a distinct peak appears around 0.35π, indicating that
the filter significantly amplifies signals at this frequency, resembling a resonator or
bandpass filter. Beyond this point, the response decreases, showing attenuation at
higher and lower frequencies. The phase response exhibits a nonlinear transition, with
a sharp shift near the resonant frequency, suggesting that the filter introduces varying
phase delays across different frequencies. This behavior is common in signal
processing applications such as audio filtering and communication systems, where
selective frequency enhancement or suppression is required.

Critical Analysis:
The lab on Z-Transform and Inverse Z-Transform provides a comprehensive
exploration of digital signal processing techniques, particularly focusing on system
analysis using transfer functions and frequency responses. The first few tasks
emphasize the partial fraction decomposition of rational functions using the
residuez function, which is essential for breaking down complex systems into simpler
components. This step is crucial for understanding how discrete-time systems
behave in the z-domain, allowing us to determine their stability and response
characteristics. By computing residues, poles, and direct terms, the lab effectively
illustrates how different terms in the expansion contribute to the overall system
behavior. These decompositions provide foundational insights for inverse Z-
transforms, enabling us to reconstruct time-domain signals from their z-domain
representations.

The latter part of the lab shifts focus to frequency response analysis using the freqz
function, which is key in evaluating how filters and systems modify input signals. By
examining the magnitude and phase responses of a given digital filter, we gain insight
into its frequency-selective properties, such as resonance and attenuation. The
observed peak in the magnitude response and the phase shift behavior demonstrate
how digital filters alter signals across different frequencies. This section reinforces the
practical significance of the Z-transform in filter design, showing how poles and zeros
influence system characteristics. The impulse response calculations in previous
tasks complement this by linking the filter's theoretical behavior to its actual time-
domain response, bridging the gap between mathematical analysis and real-world
applications.

Critically, the lab effectively combines theoretical concepts with computational

implementation, but a deeper exploration of inverse Z-transforms through direct
computation could further enhance understanding. While the partial fraction method
helps break down transfer functions, manually computing inverse Z-transforms for
specific cases would provide additional intuition on how systems respond over time.
Additionally, the frequency response analysis could be expanded by incorporating
different filter designs to compare how various pole-zero configurations affect signal
processing. Overall, the lab successfully demonstrates the power of the Z-transform
in analyzing discrete-time systems, making it a valuable exercise in digital signal
processing.