Department of Electrical and Electronic Engineering

Bangladesh University of Engineering and Technology

EEE 438: Digital Communications Laboratory

Experiment 5: Study of M-PSK and M-QAM Communication

Systems over Noisy Channel

M-ary Phase-Shift Keying (M-PSK)

In an M-ary PSK modulator, there are M possible phases that are generally equally spaced in the
constellation diagram. The incoming bit stream is divided into groups, each having log 2 M bits.
Each group of bits is forms a symbol. Each group of bits is then assigned a specific phase based
on a mapping algorithm. Thus, an M-ary PSK modulator can be defined as

௞  =  sin(2௖  + ௞ )
where ௞ =
ଶగ
ெ
for = 0,1,2, … , − 1. We can expand this signal to express it as a

summation of an in-phase (i.e. a sine carrier) and a quadrature (i.e. a cosine carrier) component.

௞  =  cos ௞ sin(2௖ ) +  sin ௞ cos(2௖ )
It is important to understand what we can call an intelligent mapping of the symbols. Figure 1
shows two possible mappings of the 4-PSK (also called QPSK) symbols. Both mappings are
equally valid from an analytic perspective, but the second one has an additional advantage. The
second map ensures that neighboring symbols differ by only one bit. Such a mapping is called
Gray Mapping. If the transmitted signal is corrupted by noise and the receiver decodes the
symbol with an error, it is most likely that the transmitted symbol will be erroneously decoded as
one of its neighboring symbols. In case of Gray mapping, such an error will cause a single bit
error and bit error rate will be minimized.
Gray Mapping for an 8-PSK system is given in Figure 2.
It should be noted that each point that represents a symbol in the constellation diagram can be
represented with a coordinate ( cos ௞ ,  sin ௞ ). The x value of the coordinate represents the
amplitude of the sine carrier and may be called the in-phase component of the signal. The y
value, on the other hand, represents the amplitude of the cosine carrier and is called the
quadrature component.

Fig. 1: Constellation of a QPSK system: (a) Binary mapping, (b) Gray Mapping

Figure 2: Constellation diagram for 8-ary PSK with Gray Mapping

Demodulation of M-PSK signal involves extracting the (noise corrupted) in-phase and
quadrature components. After passing through an AWGN channel, the received signal is given
by
 ௞  =
௞  + ௞ ()

At the receiving end, we mix the signal with the same carrier (phase matched). The resulting
signal is given by
௞  sin2 =  cos ௞ sinଶ2 +  sin ௞ sin2 cos(2௖ ) + ௞  sin2

1 1 1
=  cos ௞ −  cos ௞ cos4 +  cos ௞ sin4 + ௞  sin2
2 2 2
Taking average over a bit interval, we obtain the estimated value of the in-phase component.
௞்್
1 1
௜ =  ௞  sin2  =  cos ௞ + ௡,௜
௕ 2
ሺ௞ିଵሻ்್

Similarly, the quadrature component can be estimated as

௞்್
1 1
௤ =  ௞  cos2  =  sin ௞ + ௡,௤
௕ 2
ሺ௞ିଵሻ்್

The terms ௡,௜ and ௡,௤ appear because of the noise term, it is a random variable with a mean
value of 0 and standard deviation governed by the noise power.
We calculate the distance of the point (2௜ , 2௤ ) from all symbol coordinates
( cos ௞ ,  sin ௞ ) and choose the symbol
௞ so that this distance is minimum. Demodulation
for M-PSK systems can be given as the following block diagram.

Figure 3: Demodulation for an M-ary PSK system

Geometrically, these decision regions can be illustrated as follows.

 Figure 4: Decision regions for M-PSK

Lab Task:
(a) Write a MATLAB code that (i) implements QPSK modulation with Gray code mapping,
(ii) implements demodulation of a QPSK signal and (iii) generates the BER vs SNR plot
for the QPSK system.
(b) Observe the PSD of QPSK signal and compare its bandwidth with that of BPSK
modulated signal.
(c) Using your code from Experiment 4, generate the BER vs SNR plot for BPSK modulated
signal and compare it with that of QPSK modulated signal. Comment on your
observations

Home Task:
(a) Generalize your code from Lab Task (a) for M-PSK system.
(b) Observe the PSD of M-PSK signal for M = 2, 4, 8, 16 and 32. Comment on your
observations
(c) Plot BER vs SNR performance for M = 2, 4, 8, 16 and 32 on the same figure. Comment
on your observations.
M-ary Quadrature Amplitude Modulation (M-QAM)
M-QAM differs from M-PSK in the sense that it simultaneously encodes the symbol information
in the amplitude and phase angle of the carrier. An M-ary PSK modulator can be defined as

௞  = ௞ sin(2௖  + ௞ )
for = 0,1,2, … , − 1. We can expand this signal to express it as a summation of an in-phase
(i.e. a sine carrier) and a quadrature (i.e. a cosine carrier) component.

௞  = ௞ cos ௞ sin2௖  + ௞ sin ௞ cos2௖ 
= ௞,௜ sin2௖  + ௞,௤ sin2௖ 
There are different ways of generating the constellation of QAM signals. Typically we have
= 2௅ so each of the M symbols is L bits long. One of the obvious ways of generating a
rectangular constellation is dividing the L bits in two groups (of equal or almost equal lengths).
One of these groups modulates the amplitude of the sine carrier and comprises the in-phase
component of the QAM signal. The other group modulates the amplitude of the cosine carrier
and comprises the quadrature component of the QAM signal. The following figure illustrates
how the constellation of an 8-QAM is generated.

Figure 5: (a) Assigning amplitude levels for in-phase and quadrature bits
(b) Constellation of an 8-PSK

Demodulation of M-QAM signals essentially follows the same principal as demodulation of M-

PSK signals. One has to determine the noise corrupted in-phase and quadrature components and
then find out which symbol lies closest to it. The block diagram of M-QAM demodulator is
given below.
 Figure 6: Demodulation for an M-QAM system
Lab Task:
Write a MATLAB code that (i) implements 16-QAM modulation with Gray code mapping, (ii)
implements demodulation of a 16-QAM signal and (iii) generates the BER vs SNR plot for the
16-QAM system.

Home Task:
(a) Generalize your code from Lab Task for M-QAM system.
(b) Observe the PSD of M-QAM signal for M = 4, 8, 16 and 32. Comment on your
observations
(c) Plot BER vs SNR performance for M = 4, 8, 16 and 32 on the same figure. Comment on
your observations

Prepared by: Avik Roy

Lecturer, Department of EEE, BUET

Supervised by: Dr. Md. Saifur Rahman

Professor, Department of EEE, BUET

Based upon the previous lab manual prepared by Dr. Mohammed Imamul Hassan Bhuiyan, Md. Farhad Hossain and later
modified by Shama Naz Islam and Sajid Muhaimin Choudhury
Appendix I: Generating a Gray Map Sequence for n bit binary strings

There is a simple recursive algorithm of generating Gray map sequences. Assume that n – 1 bit
Gray map sequence is given by {ଵ , ଶ , ଷ , … ௠ } where  = 2௡ିଵ
Step 1: Write the n – 1 bit Gray map sequence followed by the entire sequence in reverse order
ଵ , ଶ , ଷ , … ௠ ௠ , ௠ିଵ , ௠ିଶ , … ଵ
Step 2: Concatenate each member of the first half of these members with a 0 and each member of
the second half of these members with a 1. This gives the Gray sequence for n bits

For example, a two bit Gray sequence is given by {00, 01, 11, 10}
After step 1, we write the sequence and its reverse order side by side
00, 01, 11, 10 10, 11, 01, 00
After step 2, we add a 0 to the right of the first half of the strings and an 1 to the right of the
second half of the strings
00, 01, 11, 10 10, 11, 01, 00
This is a Gray sequence for 3 bit binary strings