Global J. of Engng. Educ., Vol.11, No.

1
Published in Australia

2007 UICEE

Using MATLAB to Improve Learning Effectiveness and

Quality in an Undergraduate Course on Wireless
Communications and Systems
James Z. Zhang, Robert D. Adams & Kenneth Burbank
Western Carolina University
Cullowhee, North Carolina 28723, United States of America

There are two essential elements for an undergraduate engineering student to successfully master
knowledge in a course on wireless communications and systems: mathematical background including random signals and stochastic processes, and hands-on experience that allows the student to
perceive a wireless information transfer process. In the authors three-year research in teaching
communications courses, a systems approach was employed for teaching and a modular approach
used for implementing wireless communications laboratory experiments. However, while significant improvements in teaching and learning quality have been witnessed, disconnects between
theory and practice among student learners have also been observed. During the past two years,
this issue was approached by using MATLAB simulation tools. It has been demonstrated that this
approach is an excellent solution to the problem. In this article, the authors report on the details of
the design and implementation methods, how simulation is used to bridge the gap between theoretical lectures and laboratory experiences, and improvements of students learning outcomes.

INTRODUCTION

the authors conducted a three-year research on teaching methods to help students understand concepts of
electronic communications. The authors proposed
and implemented a system approach to teaching
communications fundamentals [1]. A modular
approach was utilised for teaching wireless communications [2]. A set of RF experiments using minicircuit modules was also designed and developed to
accompany a senior-level wireless communications and
systems course [3]. The goal of these efforts was to
enhance students learning experiences by enabling
students to obtain a balanced knowledge of the
desired skill sets.
The implementation of these methods has shown
significant improvements in students theoretical
understanding of wireless communications, as well as
their capabilities of hands-on skills. However, despite
this success, disconnects between theories and interpretations of practical results were observed among
the student learners. This article reports the authors
findings to use MATLAB as a vehicle to further
bridge the gap between theoretical and experiential
knowledge in the students learning processes.

Due to the nature of Radio Frequency (RF) signals

and their transmission channel characteristics, there
are two essential elements for undergraduate students
in mastering knowledge in wireless communications: a
strong mathematical background and hands-on exposure
to RF experiments. Mathematical skills include the
complex representation of signals, random variables
and stochastic processes, spectral representation and
estimation techniques, and so forth. A strong mathematical background helps students better understand
RF signal properties and transmission channels characteristics using abstract thinking that are hard to perceive
otherwise. Unfortunately, most undergraduate students
do not necessarily possess sufficient mathematics to
thoroughly understand wireless communication systems.
In this case, practical experiences greatly enhance
students understanding of wireless communication
theories. Typically, experiential exploration for students
is in the form of laboratory experiments.
In an effort to improve teaching and learning
effectiveness, and quality of communications courses,
45

46

J.Z. Zhang, R.D. Adams & K. Burbank

COURSE DESIGN RATIONALE AND

BENEFITS
As a powerful tool for mathematical analysis and
computational simulation, MATLAB has been a
predominant software package that is widely used by
educators, researchers and practicing engineers. Its
mathematical model-based inputs require users to
understand the general concepts and theories of
operation of the system to be simulated, and its excellent graphical display capability provide users with
visual simulation results that are similar to experimental
results. Its compatibility with most Digital Signal
Processing (DSP) Integrated Circuits (ICs) offers
relatively simple implementation and the realisation
of systems for feasibility studies. Because of these
advantages offered by MATLAB, it is the best tool to
be utilised for bridging the gap between theoretical
and practical understandings of a wireless communication system.
As part of the authors research, MATLAB has been
used in conjunction with course instructions for the
past three years. The focus was on illustrating technical concepts and problem-solving skills. Some research
results can be found in refs [4] and [5]. These results
have been systematically applied and expanded in a
course on wireless communications and systems by
guiding students to design and implement a wireless
system in MATLAB, and comparing the simulation
results to the theoretical results from the lectures, and
practical results from laboratory experiments. Details
of the course design and a design example are given
below.
Using a combination of lectures, laboratory experiments and MATLAB simulations, improvements in
teaching and learning quality have been observed.
In particular, the following results can be achieved
when MATLAB is used as a vehicle to enhance
students understanding between theoretical and
practical knowledge:

Improve mathematical understanding because

students are required to write mathematical modelbased scripts for simulation, they need to understand mathematical representations of each
communications block and its functionality. This
practice would greatly enhance and consolidate
students theoretical knowledge obtained from the
lectures;
Better understand practical results because the
designed laboratory experiments correspond to
simulation blocks, students can visually compare
the simulated results and those obtained from hardware in the laboratories. Students can improve

their understanding by comparing and contrasting

the results from simulations and the laboratory
experiments. Because the simulation results are
mathematical model-based, students can better
grasp the mathematics behind each functional
block for a communication system;
Bridge the gap between theory and practice as
described above, students are expected to make
better connections between theory and knowledge
from experiential exploration;
Augment system understanding because the
semester-long project is geared with lectures and
laboratories, students would have completed
a functional hardware system and a simulated
counterpart, so it is expected that students would
better understand communications concepts from
a systems perspective;
Boost simulation design capability as a bonus of
this practice, students are expected to improve
their fluency of using MATLAB as a simulation
tool. This will greatly help students with subsequent courses, such as the Digital Signal
Processing course (offered one semester after
the wireless communications course).

In the following sections, the authors report on the

design details, a design example, observed improvements in learning outcomes and future work for
further improvements.

COURSE DESIGN DETAILS

Wireless Communications and Systems (TEL 444)
is a senior-level course offered at Western Carolina
University, Cullowhee, USA. Students taking this
course should have had junior-level Electronic
Communications Fundamentals (TEL 312) as a
prerequisite. Most students also take Digital Signal
Processing (ECET 461) after taking TEL 444.
TEL 444 is a one-semester (15 weeks on average)
course. It is comprised of three major activities:
lectures, laboratories and recitation. There are four
sessions per week: two lectures, one laboratory and
one recitation. The lectures focus on the theories of
operations, mathematical representations and derivations, as well as system architectures and performances. Laboratory experiments focus on the
functionalities of each system block and measurements
of system performance parameters. MATLAB
simulation designs are closely geared with the
lectures and laboratories, and are conducted during
the recitation session. The lecture content coverage
and instruction timeline are listed in Table 1, with
the corresponding laboratories and simulation

Using MATLAB to Improve...

designs listed in Tables 2 and 3, respectively.
Part I is a review of digital modulation schemes
with an emphasis on those for wireless communications. Part V addresses wireless standards associated
with wireless networks, such as cellular systems,
Bluetooth technology and 802.11x networks. Part II
through IV is the core of this course, with the laboratories and MATLAB simulations designed to strengthen
students knowledge in these parts.
The laboratories start in the third week and end in
the twelfth week of the semester. All laboratory
experiments are designed with mini-circuits modules.
Each module can be tested individually and a 900 MHz
point-to-point wireless communication system can be
constructed by interconnecting the modules as
described in ref. [3]. Signal time domain and frequency
domain characteristics are measured using Agilent
oscilloscopes and spectrum analysers, respectively.
A comparison between Tables 1 and 2 reveals that
the designed laboratories follow the timeline of the
lectures.
Table 1: The course contents and teaching timeline.
Part
Content
I
Modulation schemes for
wireless communication
II RF communications function
blocks
III Channel characteristics and
performance
IV Wireless system performance
parameters
V Wireless system standards

Time
2 weeks
4 weeks
4 weeks
2 weeks
3 weeks

Table 2: The designed RF laboratory modules.

Lab
Description
1 I/Q Modulator and Modulated Signal
Characteristics
2 IF Filter (10.7 MHz) and Bandwidth
Control
3 I/Q Demodulator and Its Applications
4 Low Pass Filtering and Signal
Reconstruction
5 IF System Performance in an AWGN
Channel
6 Voltage Controlled Oscillator
7 RF Frequency Mixer and Frequency Upconversion
8 RF (915 MHz) Low Noise Amplifier and
Filter Characteristics
9 RF System Performance Measurements
10 Propagation and Path Loss

47

It can be observed that the simulation projects do

not exactly follow the laboratory experiments. This is
because frequency conversion only involves a simple
multiplication of the carrier that is represented by a
cosine (or sine) waveform, and the Finite Impulse
Response (FIR) filter design techniques are similar
for centre frequencies at 10.7 MHz and 900 MHz.
Designs at the IF frequency of 10.7 MHz are
sufficient for students to grasp theories and design
concepts.
One of the significant benefits of using MATLAB
simulation is to complement those aspects that are
difficult to cover with laboratory experiments. Two
major MATLAB modules are introduced to help
remove these deficiencies. The first module introduces
error correction techniques using DSP. It is not feasible to implement error correction hardware to work
with mini-circuits modules and error correction is not
part of this course. Furthermore, the limited time for
the course does not allow students to experiment with
the hardware even if it is available. On the other hand,
an error correction mechanism can be implemented
in MATLAB with relative ease and students can gain
a better understanding of how a wireless systems
overall performance can be improved in terms of
Signal to Noise Ratio (SNR) versus Bit Error Rate
(BER).
The second module introduced is system performance in a Rayleigh or Rician fading channel.
Traditionally, Rayleigh and Rician models are difficult
for students to understand because both involve the
statistical modelling of received signals. In practice or
a laboratory setting, it is almost impossible to separate
a Line-of-Sight (LOS) wave from reflected waves,
thus leaving the demonstration of Rayleigh and Rician
fading effects on a wireless channel without a
solution. To the contrary, one can easily model both
fading environments utilising MATLAB to show the
impact of either fading channel on the overall
performance of a wireless communication system.
The following section illustrates a design example
that has been used in TEL 444. The same concept
can be applied to various design projects at different
levels of complexity.

A DESIGN EXAMPLE AND LEARNING

OBJECTIVES
A typical wireless communication transmitter block
diagram and its corresponding receiver block diagram
are shown in Figures 1 and 2, respectively. The
designed laboratory experiments also follow these block
diagrams. Each block corresponds to a certain minicircuits hardware module. In the actual laboratory

48

J.Z. Zhang, R.D. Adams & K. Burbank

Figure 1: RF transmitter block diagram.

Figure 2: RF receiver block diagram.

designs in the course, the I/Q modulator and
demodulator are used for a Quadrature Phase
Keying (QPSK) signal, hence, information signal is
split into In-phase (I) and Quadrature (Q) components
before they are used to modulate the carrier.
For the sake of simplicity, the Binary Phase Shift
Keying (BPSK) modulation scheme is used to illustrate MATLAB module design concepts. Considering
that the main goal is to strengthen students understanding of wireless communication concepts, as well
as the time limitations of the course, a BPSK system
serves the purpose well. A BPSK system greatly simplifies the design of other blocks, such as filters and
demodulators.
The remainder of this section includes the design
details and results of the designed modules.
The main task of module 1 is to ask students to
generate a BPSK signal. A BPSK signal can be mathematically expressed as follows:
(1)
In (1), A is the carrier amplitude, d(t) is the information data stream in Non-Return-to-Zero (NRZ)
format assuming values {-1, 1}, and is the phase
taking values {0, }.

Upon completion of this module, students are

expected to learn the following techniques:

MATLAB commands to generate a random information data at a given data rate, including
random signal generation, signal to NRZ data
conversion and data bits generation at a predefined
sampling frequency;
Using equation (1) to create a BPSK signal at
10.7 MHz;
Investigation of the time and frequency domain
characteristics of the signal, and observation of
the signal bandwidth.

In our example, the data rate is set at 128 kbps, the

sampling rate is set at eight samples per bit and the
carrier frequency is such that a 10.7 MHz IF BPSK
signal is produced. The simulated result of this module is
shown in Figure 3. The top figure shows the original
message signal and the bottom figure shows the BPSK
modulated signal. For a clearer illustration, the plot
shown has a data rate at 1024 bps and the carrier
frequency is lowered to the same frequency.
Module 2 involves learning to design an FIR
bandpass filter centred at 10.7 MHz and use this filter

Using MATLAB to Improve...

49

Figure 3: The data and BPSK modulated waveform.

to define the bandwidth needed for transmission. The
successful completion of this module enables students
to learn the following:

Filter design technique using the Parks-McClellan

algorithm;
The bandwidth requirement for BPSK transmission;
The difference between filtfilt and filter
commands.

The results are shown in Figure 4; the top graph

shows the unfiltered signal spectrum while the bottom
graph illustrates the transmitted signal spectrum. It is
worth noting the difference between filtfilt and filter;
when using MATLAB commands, filtfilt does not
introduce additional phase shift to the signal while
filter adds extra phase shift to the transmitted signal.
Module 3 focuses on the demodulation process
of the BPSK signal. Suppose the carrier signal
generated by the receiver local oscillator is
r sc(t) = Bcos(2 f ct + ), where B is the carrier
amplitude and is the initial phase, the demodulation
process is the multiplication of the received signal in
equation (1) with rsc(t). Mathematical manipulation
shows that the demodulated signal can be expressed
as follows:
(2)

From this module, students are expected to

consolidate their knowledge in the following aspects:

Demodulation is a reverse process of modulation.

The original data signal can be reconstructed from
the modulated signal by multiplying with a local
carrier at the same frequency;
Synchronisation is important in a communication
system. When the received signal is synchronised,
in equation (2) equals zero; therefore, the maximum reconstructed message signal amplitude is
obtained (the first term in equation (2));

Figure 4: The unfiltered spectrum (top) and the

transmitted spectrum (bottom).

50

J.Z. Zhang, R.D. Adams & K. Burbank

Figure 5: Original data (top) and demodulated data (bottom).

The higher ordered term exists in the demodulated signal and needs to be filtered out before
the baseband signal can be extracted (second term
in equation (2)).

The third item of the above list naturally leads to

the next module of a lowpass filter design and signal
reconstruction. The results of module 3 are shown in
Figure 5.
Module 4 aims at the design of a lowpass filter that
removes the higher ordered term from the demodulated signal. Because the transmitted data rate is at
128 kbps, a filter with a minimum cut-off frequency at
128 kHz is needed to accomplish the task. The original

data signal and the reconstructed message signal are

shown in Figure 6.
It can be observed from Figure 6 that the reconstructed signal does not preserve its original waveform shape due to filtering. However, zero crossings
of this signal do correspond to the boundaries of
the binary data stream d(t). Consequently, a zero
crossing detector and a waveform shaper need to be
designed to completely reconstruct the original binary
data. Results of the zero crossing detection and
waveform shaper output are shown in Figure 7. Zero
crossings are indicated by the stems and the waveform shaper output is shown in bold in Figure 7
(bottom).

Figure 6: Original data (top) and low pass filtered data (bottom).

Using MATLAB to Improve...

51

Figure 7: Detected zero-crossings (top) and reconstructed binary data (bottom).

Students are expected to learn the following:

Design of a lowpass filter with the correct bandwidth;

Zero-crossing detector design;
Waveform shaper design.

Theoretically speaking, a complete wireless system

design is accomplished. The rest of the modules are
aimed at simulating the performances of the system.
These include performances in an Additive White
Gaussian Noise (AWGN) channel, a multi-path
environment and path loss of the signal.
Module 5 investigates the impact of white noise on

the signal transmission. White noise is added to the

received signal before demodulation. The reconstructed signal is shown in Figure 8. Compared to its
non-noisy counterpart in Figure 6 (bottom), it can be
observed that zero-crossings are not preserved due to
noise contamination of the received signal. This means
that errors occur in the received signal. Different
levels of noise are added to the received signal to
allow students to observe the severity of signal distortion. Students start to appreciate a clean transmission
environment and realise the hostilities of a wireless
transmission channel.
Module 6 is a natural extension of module 5. In this
module, students are required to assess the systems

Figure 8: Original data (top) and received data through an AWGN (bottom).

52

J.Z. Zhang, R.D. Adams & K. Burbank

performance by calculating BERs that correspond to

various SNRs. Figure 9 shows the results of this simulation. Through this module, students gain a better
understanding of the system performance in an AWGN
channel. It is worth noting that the plot is BER versus
SNR rather than traditional BER versus Eb/N0.
Yet another factor that severely decreases a
wireless systems performance is multipath. Module
7 addresses this concern by requiring students to
simulate Rayleigh and Rician fading channels. Rician
fading contains a dominant LOS signal plus a collection of infinite number of reflected waves and can be
mathematically modelled as follows:
(3)
In (3), An and n are the amplitude and phase
of the nth ray, respectively. n normally depicts a

Figure 9: System performance: SNR versus BER.

delayed arrival of signals.

Rayleigh fading is a collection of infinite number
of reflected waves without an LOS. Therefore, by
setting A = 0 in equation (3), a Rayleigh fading
channel is recovered.
Even though the theory calls for an infinite
number of rays, it is impossible to simulate this
condition. For the purpose of enhancing students
learning, 3~5 rays are sufficient. When a 3-ray model
is used, the recommended values for An and n are
A/8, A/4 and A/2, and , /2 and /4, respectively.
Figure 10 shows the received Rayleigh and Rician
signals. When compared to the original transmitted
signal in Figure 3, a distortion of the signal can be
observed.
Students are asked to carry out the same tasks in
module 6 where the systems performance is assessed.
Students will observe that, unlike that in an AWGN
channel, the systems performance in a fading
channel improves slowly with increased SNR.
As described in the previous section, it is not
feasible to implement error detection and correction
mechanisms given the current equipment available.
In this case, MATLAB provides an excellent opportunity for students to appreciate the power of DSP
techniques in detecting and correcting errors. Although
the detection and correction method is very primitive,
the concept is well conveyed to the students.
Module 8 guides students to carry out a task of
detecting and correcting errors in the reconstructed
signal. Because the message data is over-sampled at
eight samples per bit, therefore, any reconstructed
bit that does not contain exact eight consecutive 1s or
-1s is an error bit. The majority rule for the decision is

Figure 10: Received signals in Rician and Rayleigh channels.

Using MATLAB to Improve...

53

used in this brutal force approach: the MATLAB

program checks bit-by-bit by checking eight sampling
points at a time. If the majority of these sampling points
have values of 1, other sample values are also forced
to be 1. Otherwise, a decision is made that the
transmitted bit is a -1. Although somewhat timeconsuming when the transmitted data stream is of a
large size, nevertheless, the detection and correction
job is well done. Uncorrected and corrected received
data are shown in Figure 11.
The last module is designed to investigate the
propagation properties of the signal and pass loss. The
theoretical received signal strength is mathematically
expressed as follows:
(4)
where Pr(d) is the received signal power at distance
d, Pt is the transmitted power, Gt and Gr are transmitter and receiver antenna gains, respectively, is the
signal wavelength, and d is the distance between the
transmitter and the receiver.
Because the path loss is reflected by the received
signal amplitude reduction, distance cannot be
directly simulated in MATLAB. Therefore, the
strategy is changed slightly to achieve the maximum
learning outcome. In the module, students are required
to measure the overall received signal strength
and calculate the path loss of the system. This is
accomplished by laboratory module 10 listed in Table
2. The details of laboratory procedures can be found
in ref. [3].
Figure 12 is a plot of the measured received signal
power from the laboratory and theoretical received
signal power calculated using equation (4). Path loss

Figure 12: Measured versus theoretical path-loss

comparison.
can be calculated by taking the difference between
the transmitted and received powers.

CONCLUSIONS AND FUTURE WORK

In this article, the authors report on the progress of a
three-year research project aimed at improving the
learning effectiveness and quality in a senior-level
course on wireless communications and systems.
MATLAB is used as a vehicle to enhance students learning
experiences by bridging students understandings of the
relevant theories and abilities of interpreting experimental
results. As a result, improvements have been observed
in students problem-solving capabilities.
Further research will be focused on more complicated system designs and a systematic assessment of
students learning outcomes.

REFERENCES

Figure 11: SNR versus BER with and without error

correction.

1. Zhang, J.Z., Burbank, K. and Adams, R.,

A systems approach to teaching introduction to
electronic communications for ECET students.
Proc. 2004 ASEE Annual Conf. & Expo., Salt
Lake City, USA (2004).
2. Zhang, J.Z., Burbank, K. and Adams, R.,
A modular approach to teaching wireless
communications and systems for ECET students.
Proc. 2005 ASEE Annual Conf. & Expo.,
Portland, USA (2005).
3. Zhang, J.Z., Adams, R.A. and Burbank, K., The
design of RF labs using Mini-Circuit modules to
improve the quality of teaching in a course on wireless communications and systems. World Trans.
on Engng. and Technology Educ., 4, 1, 39-42
(2005).

54

J.Z. Zhang, R.D. Adams & K. Burbank

BIOGRAPHIES

Robert D. Adams received

a PhD in electrical engineering from the University of
Alabama in Huntsville. He
has been a faculty member in
the Kimmel School of Western Carolina University since
2003. His research interests
include numerical solutions to
biomedical applications of
electromagnetics, 3D modelling and visualisation, and digital image processing.

James Z. Zhang is an assistant professor and the

Program Director of the
Master of Science in Technology Program at Western
Carolina University. He holds
a PhD (2002) in electrical
and computer engineering
from Purdue University.
His research interests
include wireless systems and
wireless networks. He is a member of the ASEE and
a senior member of the IEEE.

Kenneth Burbank received

the ScB (1973), ScM (1975)
and PhD (1979) in electrical
engineering from Brown
University, Providence. He
is the Program Director for
Electrical Engineering at
Western Carolina University, where he has been an
associate professor since
2001. Dr Burbank is active
with the IEEE and TAC of ABET, and strives to bring
practical engineering activities into the classroom.

4. Zhang, J.Z., Adams, R. and Burbank, K.,

Theory, practice, and systems a new approach
to teaching electronic communications with
Matlab. Inter. J. of Engng. Educ., 21, 3, 612-617
(2005).
5. Zhang, J.Z. and Burbank, K., Bridging the gap
between theory and practice an example of
teaching communications courses using Matlab.
Proc. 2004 ASEE-Pacific South West Conf.,
Stockton, USA, 104-115 (2004).