See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/321176321

Performance Analysis, Characteristics, and Simulation of Digital QAM

Article  in  European Journal of Electrical Engineering and Computer Science · September 2017

DOI: 10.24018/ejece.2017.1.1.3

CITATIONS READS

2 2,681

1 author:

Nikolaos F. Voudoukis
National Technical University of Athens
33 PUBLICATIONS   35 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Teaching Models of Light and Properties of Light View project

Modelling of Electronic Circuits and Applications View project

All content following this page was uploaded by Nikolaos F. Voudoukis on 07 December 2017.

The user has requested enhancement of the downloaded file.

 EJECE, European Journal of Electrical and Computer Engineering
Vol. 1, No. 1, October2017

Performance Analysis, Characteristics, and Simulation of

Digital QAM
Nikolaos F. Voudoukis

channel system, using Amplitude Shift Keying (ASK) on

Abstract—Quadrature Amplitude Modulation or QAM is a both channels. This results to a signal equivalent to one that
form of modulation which is widely used for modulating data combines PSK and ASK.
signals onto a carrier used for radio communications. QAM, There are alternative implementations of the QAM
when used for digital transmission for radio communications
modulation. The most popular types of QAM include 16, 32,
applications is able to carry higher data rates than ordinary
amplitude modulated schemes and phase modulated schemes. 64, 256 QAM. Different forms of QAM such as rectangular,
This paper presents the various fields where QAM can be circular etc. are available. We convert analog modulation to
implemented, describes modulator/demodulator block digital in order to achieve extra capacity, more secure data,
diagrams for transmitters as well as receivers, provides an higher quality of communications, better compatibility with
introduction of certain performance indicators of modulation digital data services and quicker system availability. The
and a list of applications using alternative implementations of
most usual issues for researchers and developers of
QAM. Also the paper presents a simulation of QAM using
Simulink (example of 16-QAM signal) with signal trajectors communicational systems are the available bandwidth, the
and constellation plots. It is compared the theoretical and congenital noise level that maximum admissible power and
simulated Bit Error Rate (BER) for 16-QAM with Gray coding the fact that there is only on certain R/F spectrum to be
in an AWGN channel. Some general conclusions are also cited. distributed equally to all users. That last issue is the main
constraint, as more and more users claim to share the
Index Terms— Quadrature Amplitude Modulation (QAM) available spectrum, causing great increase in the demand of
Constellation diagram, Modulator/Demodulator, Simulation,
Bit error rate (BER), Signal to noise ratio (SNR). communications services. Digital modulation patterns have
higher ability to carry larger amounts of information than
I. INTRODUCTION analog modulation schemes. In the QAM we use a confined
number of amplitudes and phases. By confined we mean at
Quadrature amplitude modulation (QAM) is both an
least two [2].
analog and a digital modulation scheme. It conveys two
Constellation diagrams consist of constellation points,
analog message signals, or two digital bit streams, by
placed in a rectangular grid with equal space both vertically
changing (modulating) the amplitudes of two carrier waves,
and horizontally. The possible states of a particular
using the amplitude-shift keying (ASK) in digital
configuration are noted very reliably on a constellation
modulation scheme or amplitude modulation (AM) in
diagram. As we have binary data in digital communication,
analog modulation scheme. QAM is a common choice in
the points on the diagram are often the result of successional
digital data modulation. The procedure involves the
multiplications by 2 (2, 4, 8, 16, 32, etc.). The most common
modulation of an analog carrier signal by a discrete signal.
ones are 16, 64, 128 and 256 QAM [1].
This means that during modulation, digital is converted to
The different positions for the states in different forms of
analog, while during corresponding demodulation, analog is
QAM are represented at constellation diagrams. The number
converted to digital. There is a finite number of M
of points on the QAM constellation diagram increases as the
alternative symbols that is used to represent the changes in
order of the modulation increases.
the carrier signal.
Square constellations of QAM do not provide optimal
The use of Quadrature Amplitude Modulation basically
reflection, as there cannot be maximum spacing of the
consists of the combination of two amplitude modulated
points for a given energy. On the opposite side, they present
signals into the same channel. QAM doubles the amount of
the advantage of being easily received as two burst signals
the bandwidth, thus making it more effective. This
which have been amplitude modulated by quadrature
combination of analog and digital modulation is very often
conveyers and they can be demodulated very easy. From the
encountered in digital telecommunication systems. Also, in
other side, non-rectangular diagrams offer a more reliable
wireless applications using pulse amplitude modulation.
bit-error rate, but aren’t as easy to modulate and
QAM is the path used to encode digital channels and send
demodulate.
them through cable television providers [1].
In theory, the possibility of using QAM with higher order
QAM consists of two signals, an anaphase one called the
for transferring bigger number of bits / symbol (higher order
I signal and a quadrature phase one, known as the Q signal.
constellations) is a solid one. But, in practice, this would
Through amplitude modulation, with a specific number of
arise a reliability issue. Higher order constellation means
amplitudes, the signals are summed and they form a two
constellation points very close to one another, if we intend
to maintain the mean energy of the system at the same level.
Published on October 23, 2017.
Nikolaos F. Voudoukis is with Department of Electrical and Electronic That kind of configuration concludes in extra noise and
Engineering Educators, School of Pedagogical and Technological distortion. So, while higher order QAM indeed carries more
Education (ASPETE), Athens, Greece (e-mail: [email protected]).

DOI: http://dx.doi.org/10.24018/ejece.2017.1.1.3 1
 EJECE, European Journal of Electrical and Computer Engineering
Vol. 1, No. 1, October2017

data, it in facts delivers the data with a higher fit error rate, transmission is small, the point that corresponds to the
thus causing it to be less reliable than lower order QAM. received signal looks like a small spot around the point that
At the figure below there are constellation diagrams for corresponds to the intended signal. In the case that we have
various formats of modulation: big noise, the point of the received signal may be found any
were (and not in a specific area) in the diagram. If someone
wants to take the information delivered from the received
signal first they have to clarify the point of the indented
signal in the diagram and after that to encode from same
point the information. There are several algorithms doing
this job, that they have been developed by scientist and
engineers for identifying the intended signal point.
Estimating the shortest possible distance between the
perceived signal and one of the constellation points leads the
algorithm to recognition of the most possible point intended
by the received signal.
The estimation of the distance is a process that takes time
and generates delays. There are few algorithms that face this
problem with a satisfactory way. For the shortest distance
between points might be developed a (performance) pattern
based on a combinational logic. This pattern must be as
small as possible in area, so the small dimensions in silicon
can achieve maximum speed and performance. Then QAM
receiver can be prosecuted with high-velocity and less
packaging consistency circuitry to reduce the coded
information at highest rate possible. QAM achieves a greater
Fig.1. QAM with different bits per symbol. distance between constellation points.
Generally, there is a non-quadratic constellation that is
Here the diagrams reveal the number of perspicuous best used for a specific. But this non-quadratic constellation
conditions that can exist represented by the same number of is not often implemented because the quadratic QAM is
points on the constellation. Most of the sequences of much easier to modulate and demodulate [1].
constellations can be configured in various ways due to the
nature of QAM. II. ADVANTAGES AND CHARACTERISTICS OF QAM
Rectangular quadrature amplitude modulations are easier QAM is a form of modulation that has many advantages:
to modulate and demodulate than non-rectangular Bit Packaging Ratio is increased, Bandwidth efficiency is
quadrature amplitude modulations. So rectangular QAM is increased, Overall channel capacity is increased, More
preferable to non-rectangular QAM [2]. number of customers can be catered, The trade-off between
Quadrature amplitude modulation technique is basically bandwidth and signal to noise analogy can be made i.e.
used in digital communication for coded signal same channel capacity is achievable at a lower signal to
transmission. The QAM technique uses two orthogonal noise ratio value because of higher bandwidth, Less
signals. These two signals with appropriate scaling will form transmission power is being required by the transmitter,
them into variant values of amplitude and phase. The waves Power budget of both the transmitter and receiver reduces
are depicted in a graphical way as points, in the diagram. by many folds thus reducing the overall cost of the system.
With this method we have the name of the QAM as 16, 32... In a telecommunication link the data rate (bits/sec) can be
256 state QAM. Cosine wave and sine wave modulation increased by the selection of a QAM format with higher
factors are in fact the coordinates (Cartesian) in the order than the already existing QAM format. [1], [2].
constellation diagram. They represent the quality and At the following table we have an overview of the
distortion of a received signal. A square lattice of 16 relation between symbol rate and bit rate of different
message points consists the signal constellation for 16-state formats of QAM.
QAM. The quadrature and the in-phase components are
modified in a way that the conveyer wave undergoes TABLE I: BIT RATES OF DIFFERENT FORMS OF QAM
amplitude modulation as well as phase modulations. One Bits per symbol Symbol Rate Modulation
signal may be attenuated while the other signal gets 1 1 x bit rate BPSK
magnified. As a result of this the on demodulation scaling 2 1/2 bit rate QPSK
3 1/3 bit rate 8PSK
factors obtained will describe in a good way the distortions
4 1/4 bit rate 16QAM
which may be obtained in the amplitude in and in the phase
5 1/5 bit rate 32QAM
angle at the two signals (in phase signal and quadrature
6 1/6 bit rate 64QAM
signal). So the point in the constellation diagram tends to
shift in a definite position. The noise which will be added to
the signals can be measured by the distance between the Much faster data rates and higher levels of spectral
points of the received signal and the points of the efficiency can be offered by higher order modulation rates
transmitted (intended) signal. If the noise influenced during for the radio communications systems. This benefit comes at

DOI: http://dx.doi.org/10.24018/ejece.2017.1.1.3 2
 EJECE, European Journal of Electrical and Computer Engineering
Vol. 1, No. 1, October2017

a price. The communicational patterns with higher order

modulation techniques are not as resistant to noise and The following picture shows the QAM Modulator
insertion. Simulink subsystem of 16-QAM digital communication
This has resulted to increased use of dynamic adaptive system [3].
formation techniques by radio communications systems.
That means that they adapt the modulation scheme after
sensing the channel conditions in order to achieve the
highest data rate for the specific status. As SNR (signal to
noise ratio) increase, error rate will decrease. On reverse, as
SNR decrease, error rate will increase. The variation of SNR
has as a result the variation of system throughput (increasing
or slowing it). The reliability of the link can be restored only
with the conversion to a lower order modulation scheme [2].

III. DESCRIPTION OF THE QAM

MODULATOR/DEMODULATOR
Fig.3. QAM Modulator
All telecommunication systems, regardless whether they
use digital or analog modulation, are made up of a number The receiver simply realizes the converse procedure of
of modules (amplifiers, filters etc.). Transmitter and receiver the transmitter.
are the most significant devices with a structure that relies The following picture shows the IQ correlation Receiver
on the modulation pattern. Here we have QAM. Simulink subsystem of 16-QAM digital communication
The digital modulation scheme (just the same as the system [3].
analog modulation scheme as well) demands a demodulator
in the receiver block that recovers the message (information
signal) from the received signal. The demodulator makes the
reverse job compared with modulator (take apart – split,
information signal and carrier).
The stream of binary digits to be transmitted is divided
into two equal parts. So, in this way we have two
congregational signals to be conveyed. These two signals
are encoded separately. The way is the same as the ASK
modulation. After that one signal (in channel that is called in
phase) is multiplied by a cosine. The other signal (in channel Fig.4. 16-QAM IQ correlation Receiver
that is called quadrature) is multiplied by a sine. As between
cosine and sine there is a phase of 90° the same phase there The following picture shows the 16-Level Bit to IQ
is between the two signals. At the end the two signals are Symbol Simulink subsystem of 16-QAM digital
added. The final signal (sum of the in phase and quadrature) communication system [3].
is sent through the real channel [1].
It is possible to extract the component in phase (or in
quadrature), if we multiply by a cosine (or a sine) and by a
low-pass filter (LPF). So there is only an ASK demodulator
and the two flows of data (bits) are merged back.
There is an undefined phase delay between the transmitter
and receiver. This phase delay must be indemnified by
synchronization of the receiver’s local oscillator; (i.e., the Fig.5. 16-Level Bit to IQ Symbol
sine and cosine functions). In mobile applications, due to the
potential presence of a Doppler shift proportional to the Both the frequency and phase differentiations imported by
respective velocity of the transmitter and receiver, there will the channel must be counterbalanced. We can accomplish
be a counterbalance in the respective frequency. that by adequately adjusting the sine and cosine elements.
The following picture shows the bit-to-symbol Simulink The above procedure demands a phase reference, and is
subsystem of 16-QAM digital communication system [3]. achieved by using a Phase-Locked Loop (PLL) [4].

IV. SIMULATION - RESULTS

In case of 16-QAM, the four bits in each point at the
constellation diagram (Gray coded) can be considered as
two pairs of two bits each on 4-PAM modulation
independently, on I-axis and Q-axis respectively.
In digital communication systems, the receiver’s basic job
is to select precisely the transmitted message symbols out of
a finite set. The fact that there is noise in the communication
Fig.2. The bit-to-symbol subsystem

DOI: http://dx.doi.org/10.24018/ejece.2017.1.1.3 3
 EJECE, European Journal of Electrical and Computer Engineering
Vol. 1, No. 1, October2017

channel makes the task more difficult and causes bit-errors.

In a digital communication system, one significant
performance criterion is the Bit Error Rate (BER). BER is
the proportion of the number of wrongfully received bits to
the total number of transmitted bits. Signal-to-Noise Ratio
(SNR) is the ratio (energy per bit)/ (noise power spectral
density). When SNR increased, BER decreased (for the
same QAM format, eg 16-QAM).
We will compare theoretical and simulated Bit Error Rate
(BER) for 16-QAM with Gray Coding in an AWGN channel
[3].
We can measure the 16-QAM BER (numerically) by
computer simulation and can verify our analytical results.
For 16-QAM modulation we can see that amplitude
estimation error gives 1-dB degradation in average SNR
(signal-to-noise ratio) and combined amplitude-phase
Fig.8. 16 QAM constellation plot with Gray coded mapping (another
estimation error gives 2.5-dB degradation for the parameters
example)
we consider [4].
In wireless communications, the bit-error rate (BER)
TABLE II: 16-QAM GRAY-CODED CONSTELLATION MAPPING
representation of M-QAM is significantly degraded by the
b0b1 I b2b3 Q
00 -3 00 -3 severe amplitude and phase variances (congenital to wireless
01 -1 01 -1 channels). This happens as the demodulator needs to scale
11 +1 11 +1 the received signal in order to regularize channel utility. So
10 +3 10 +3 its estimation districts correspond to the transmitted signal
constellation. This scaling process, which is requisite, is
known as automatic gain control (AGC) [4]. The AGC
wrongfully scales the received signal, if the channel utility is
estimated in error, which can cause false demodulation even
without the presence of noise. In order for the
communication with M-QAM to be reliable (e.g. 16, 32, 64
etc.) we need accurate fading compensation techniques at
the receiver. A truly efficient technique in M-QAM
demodulation, to counterbalance channel amplitude and
phase distortion, is channel sounding.

Example:
The following picture shows a 16-QAM signal with
amplitude A=5V peak, fc=20 KHz, Ts=4msec for the data
bits bi =1100001111.
Fig.6. Gray coded bit mapping for 16 QAM

The BER of QAM can be calculated through Monte Carlo

simulations. However, this becomes quite complex as the
constellation size of the modulation schemes increases.
Therefore, a theoretical approach is sometimes preferred.

Fig.9. 16-QAM signal

The following pictures show Signal Trajectory and

Constellation Plot for a 16-QAM signal with 4bits/symbol
and for five different values of Eb/N0 (4, 8, 12, 16, 20 dB)

Fig.7. Bit Error Rate plot for 16QAM modulation with Gray mapping

DOI: http://dx.doi.org/10.24018/ejece.2017.1.1.3 4
 EJECE, European Journal of Electrical and Computer Engineering
Vol. 1, No. 1, October2017

Fig.10. Signal Trajectory and Constellation Plot for Eb/N0 = 4dB Fig.12. Signal Trajectory and Constellation Plot for Eb/N0 = 12dB

Fig.11. Signal Trajectory and Constellation Plot for Eb/N0 = 8dB Fig. 13. Signal Trajectory and Constellation Plot for Eb/N0 = 16dB

DOI: http://dx.doi.org/10.24018/ejece.2017.1.1.3 5
 EJECE, European Journal of Electrical and Computer Engineering
Vol. 1, No. 1, October2017

When a radio Tx amplifies phase or frequency modulated

signal, there is no need to use linear amplifiers. Linearity
must be maintained when using QAM because it includes
amplitude components. However, since linear amplifiers are
inefficient and use up more power, they have become less
popular for use on mobile applications.
Link conditions and requirements such as BER, SNR,
data rate etc. are responsible for the choice of the
modulation pattern in a radio communication system. QAM
modulation is a method for reducing the bandwidth which is
needed for transmission of digital data. In case of QAM the
exchange between bandwidth and power is more beneficial
than in M-PSK.
As mentioned, for each symbol both phase and
amplitudes are varied to represent different bits. There are
two levels of amplitudes for each phase d1 and d2. There are
many variants to this technique. Most popular are 16-QAM,
64-QAM and 256-QAM. In 16-QAM each symbol
represents 4 bits as mentioned in the constellation diagram.
In 64-QAM, each symbol is represented by 6 bits and in
256-QAM, each symbol is represented by 8 bits. As it
increases in level QAM technique becomes more bandwidth
efficient but it requires very robust algorithms to decode
complex symbol at receiver into bits. For example, 256-
QAM is complex than 16-QAM.
QAM is less robust but it is more bandwidth efficient
compared to BPSK. Thus for poor CINR (Carrier to
Interference and Noise Ratio) BPSK is employed, but for
better CINR, QAM is employed, which leads to better data
Fig.14. Signal Trajectory and Constellation Plot for Eb/N0 = 20dB rate. Numerous radio communications and data delivery
applications use QAM. Despite this, a few particular
From the above pictures we can see that as Eb/N0 variations of QAM are implemented in specific applications
increased (from 4 to 20 dB, in our example) the Signal and standards. In example, for domestic broadcast
Trajectories are clearer and the Constellation Plots have the applications 64 QAM and 256 QAM are often encountered
points in a more specific area (small error). So if we in digital cable television and cable modem applications. In
consume (give) more power we have better the US, 64-QAM and 256-QAM are the binding modulation
telecommunication link. patterns for digital cable as standardized by the SCTE in the
standard ANSI/SCTE 07 2000. In Great Britain, 16 QAM
V. DISCUSSION and 64 QAM are presently applied in digital terrestrial
Digital modulation, like analog modulation, changes a television using DVB (Digital Video Broadcasting). Other
transmittable signal (carrier) based on the information types of QAM are also applied in various cellular and
received by a message signal (information signal). wireless technology applications [4]. Direct modulation of a
In digital transmission applications [1], QAM allows laser has always been used in optical telecommunications.
higher data rates than usual phase modulated and amplitude Although distance–capacity product capabilities using this
modulated patterns. The number of points on the approach have been limited, it represents a very cost-
constellation diagram, which are the points at which the effective solution, making it a prime choice especially when
signal can rest, as with phase shift keying (PSK), is outlined transmission distance is limited, for example, within a data
in the modulation format display, for example 64-QAM uses center or a supercomputer (hundreds of thousands within a
a 64-point constellation. The most usual types of QAM use a single supercomputer) or in ‘last-mile’ telecommunications.
number of points on the constellation diagram which is [5], [6].
equal to a multiplication by 2 such as, 2, 4, 8, 16 etc. Application examples: CDMA WiMAX-16d, 16e,
because QAM constellation points are usually cited in a WLAN-11, an OFDM Satellite DVB, Cable modem.
plexus with even vertical and horizontal spacing. Using Equipment like videophones, mobiles and in numerous
higher order modulation formats enables transmitting more defense applications often uses large scale circuit integration
bits per symbol, but presents the disadvantage of being more schemes based on QAM technique.
vulnerable to noise and data errors as the points are closer
together. VI. CONCLUSION
Drawbacks of QAM are: a. Susceptibility to noise is From the simulation we can conclude (expected from the
higher due to the closeness of the states. The slightest noise theory) that as Eb/N0 increased (so SNR increased), BER
is sufficient to cause movement of the signal to an reduced. Interference typically increases when we are
alternative decision point and provoke data corruption. b. moving to a higher order QAM constellation (more

DOI: http://dx.doi.org/10.24018/ejece.2017.1.1.3 6
 EJECE, European Journal of Electrical and Computer Engineering
Vol. 1, No. 1, October2017

constellation points, higher data rate and mode) in hostile [2] B. Sklar, Digital Communications - Fundamentals and Applications,
Prentice Hall, New Jersey, 2001.
environments, such as RF/microwave, wireless, cellular [3] D. Silage, Digital Communications Systems using Matlab and
QAM applications (e.g. in broadcasting or cellular Simulink, Bookstand Publishing, 2009.
telecommunications multipath). There is reduced noise [4] W. T. Webb and L. Hanzo, Modern Quadrature Amplitude
Modulation, New York, IEEE Press, 1994.
immunity because of a spreading of the spots in the
[5] Schindler P. C. et al. “Monolithic GaAs electro-optic IQ modulator
constellation, which as a result gives a decreasing of the demonstrated at 150 Gbit/s with 64QAM.” J. Lightwave Technol. 32,
separation between adjacent states, thus making it difficult 760–765, 2014
for the receiver to decode the signal appropriately. [6] Chiba A. et al. “16-level QAM by monolithic quad-parallel Mach-
Zehnder optical modulator”, Elctron. Lett. 46, 220–221, 2010.
Estimation of signal to noise ratio for the received signal is
an important task in communication systems. The Nikolaos F. Voudoukis received a BSc degree in
knowledge of the SNR is a requirement on many systems in Physics from Athens National University, Greece, in
1991, a BSc in Electrical and Computer Engineering
order to perform efficient signal detection and link from the National Technical University of Athens,
adaptation. We are able to determine an optimal QAM mode Greece, in 2012, his MSc degree in Electronics and
for a specific operating environment because there are Telecommunications from Athens National
University, in 1993, and his PhD degree from Athens
several test parameter measurements to support the best National University, in 2013. He has worked as
solution, for a given application. telecommunication engineer in Greece. Dr
Voudoukis now is Assistant Director at a high school
and a part-time Lecturer at the School of Pedagogical & Technological
REFERENCES Education, Athens, Greece.
[1] J. G. Proakis, Digital Communications, New York, NY, McGraw-
Hill, 2001.

DOI: http://dx.doi.org/10.24018/ejece.2017.1.1.3 7

View publication stats