Performance Analysis of Adaptive Modulation and Coding Schemes on OFDMA Physical Layer

TL;DR: Performance evaluation of Adaptive Modulation and Coding (AMC) in downlink of an orthogonal frequency division multiple access (OFDMA) network, considering Partial Usage of Sub-channels (PUSC) suggests better performance for AMC over individual MCS in all channel environments.

Abstract: 3 Abstract: The recent demand for higher data rate services from wireless network users is overwhelming. Social media influx as well as the proliferation of broadband enabled smart-phones, tablet computers and other newly improved wireless devices has erupted a new trend in wireless network traffic need where average capacity and speed is no longer appreciable. In order to cope with this trend in traffic requirement, wireless network operators are considering a gradual rollover of an existing third generation (3G) network to a fourth generation (4G) network with orthogonal frequency division multiple access (OFDMA) based technologies such as Fourth Generation Long Term Evolution (4G LTE) and Worldwide Interoperability for Microwave Access (WiMAX). This paper is devoted to the performance evaluation of Adaptive Modulation and Coding (AMC) in downlink of an orthogonal frequency division multiple access (OFDMA) network, considering Partial Usage of Sub-channels (PUSC). By using MATLAB Simulink and Origin 61, the performance of Bit Error Rate (BER) and Spectral Efficiency in two channel environments, i.e. non-fading and fading channels were examined. Results suggest better performance for AMC over individual MCS in all channel environments. Moreover, non-fading Addictive White Gaussian Noise (AWGN) channels significantly perform better than fading (Rayleigh and Rician) channels. In Rician channel environment, however, flat fading Rician channels perform better than frequency selective Rician channel which interestingly records a degraded performance against Rayleigh channels. Keywords: third generation (3G), fourth generation (4G), Orthogonal Frequency Division Multiple Access (OFDMA), Long Term Evolution (LTE), Worldwide Interoperability for Microwave Access (WiMAX), Bit Error Rate (BER), Adaptive Modulation and Coding (AMC), Partial Usage of Sub-channels (PUSC), Rayleigh channels and Rician channels

Full text

IJARCCE
ISSN (Online) 2278-1021
ISSN (Print) 2319 5940
International Journal of Advanced Research in Computer and Communication Engineering
Vol. 4, Issue 12, December 2015
Copyright to IJARCCE DOI 10.17148/IJARCCE.2015.41210 44
Performance Analysis of Adaptive Modulation
and Coding Schemes on OFDMA Physical Layer
Sayawu Yakubu Diaba
1
, Theophilus Anafo
2
, Zuberu Abdul-Rahman
3
Suhum Gariba Zongo, Ghana
1
Tema, Ghana
2
Rautkalliontie 4E 01360 Vantaa, Finland
3
Abstract: The recent demand for higher data rate services from wireless network users is overwhelming. Social media
influx as well as the proliferation of broadband enabled smart-phones, tablet computers and other newly improved
wireless devices has erupted a new trend in wireless network traffic need where average capacity and speed is no longer
appreciable. In order to cope with this trend in traffic requirement, wireless network operators are considering a gradual
rollover of an existing third generation (3G) network to a fourth generation (4G) network with orthogonal frequency
division multiple access (OFDMA) based technologies such as Fourth Generation Long Term Evolution (4G LTE) and
Worldwide Interoperability for Microwave Access (WiMAX). This paper is devoted to the performance evaluation of
Adaptive Modulation and Coding (AMC) in downlink of an orthogonal frequency division multiple access (OFDMA)
network, considering Partial Usage of Sub-channels (PUSC). By using MATLAB Simulink and Origin 61, the
performance of Bit Error Rate (BER) and Spectral Efficiency in two channel environments, i.e. non-fading and fading
channels were examined. Results suggest better performance for AMC over individual MCS in all channel
environments. Moreover, non-fading Addictive White Gaussian Noise (AWGN) channels significantly perform better
than fading (Rayleigh and Rician) channels. In Rician channel environment, however, flat fading Rician channels
perform better than frequency selective Rician channel which interestingly records a degraded performance against
Rayleigh channels.
Keywords: third generation (3G), fourth generation (4G), Orthogonal Frequency Division Multiple Access (OFDMA),
Long Term Evolution (LTE), Worldwide Interoperability for Microwave Access (WiMAX), Bit Error Rate (BER),
Adaptive Modulation and Coding (AMC), Partial Usage of Sub-channels (PUSC), Rayleigh channels and Rician
channels
I. INTRODUCTION
There has been enormous growth in the
Telecommunication and ICT industry over the past few
decades with mobile and wireless communication
dominating. Presently, there are over six billion eight
hundred million mobile cellular subscribers globally with
mobile-broadband subscriptions approaching two billion
three hundred million. [1,2]. The integration of voice,
video and data, which is providing multimedia services
has given rise to the demand for the progressive drift from
high capacity voice services to high speed and high
capacity data services.
The arrival of sophisticated mobile devices and the
tremendous bandwidth requirement of certain services and
applications such as video calls, video conferencing,
online games and the increased utilization of data websites
such as Facebook and You tube ignite the need for higher
capacity and higher data rate supported technology [3, 17].
Hence the development of the 4G (4th generation)
Orthogonal Frequency division multiple access (OFDMA)
based wireless technologies such as Mobile WiMAX
(Worldwide Interoperability for Microwave Access) and
3GPP (3rd Generation Partnership Project) LTE (Long
Term Evolution) which have been identified as promising,
permitting higher data rate in wireless broadband access
[5]. 4G networks such as WiMAX, LTE and UMB which
are currently in early deployment stage in some and yet to
be deployed in some offer speed from 100Mbps to 1Gbps
providing enormous data rate for bandwidth intensive
services and applications.
The concept of WiMAX is based on the Open Systems
Interconnections (OSI) reference model which has the
lowest layer as the physical layer which is based on
OFDM technology. It is responsible for the specification
of the bandwidth, modulation and coding scheme, data
rate, multiplexing, error correction, transmitting data in
frames, synchronization between transmitter and receiver
and controlling access to shared wireless medium
classified under media access control (MAC) layer [4].
The IEEE 802.16 (WiMAX) standard is as a result of the
increasing higher capacity and higher speed
communication required for video, voice and multimedia
to meet the growing effect on how people interact or
communicate as well as enjoy their entertainment.
WiMAX standards have evolved through the years since
its introduction in 2001 with its IEEE 802.16 standard
which was also known as fixed WiMAX [8].
Interoperability is the key objective of WiMAX which
ensures that, equipment from different vendors
interoperates in the framework without difficulty.

IJARCCE
ISSN (Online) 2278-1021
ISSN (Print) 2319 5940
International Journal of Advanced Research in Computer and Communication Engineering
Vol. 4, Issue 12, December 2015
Copyright to IJARCCE DOI 10.17148/IJARCCE.2015.41210 45
International standards do not define boundaries for
subscriber module usage. It supports scalable OFDMA,
QPSK, 16QAM and 64QAM with channel bandwidth
selection within the range of 1.75MHz to 20MHz [5, 7].
However, signal power loss is prevalent along the
propagation path for certain frequencies hence leading to
limited coverage. Additionally, dead spot and areas of
poor reception caused due to fading and blocking within
coverage further reduce the efficiency of this standard.
Conversely, one of the salient efforts to solve this problem
has been to deploy relatively more base stations (BS)
within a geographical area [13]. Resource allocation for
both uplink and downlink are controlled by a scheduler in
the base station.
The authors of [10,11] stated that, the WiMAX physical-
layer is capable of supporting MIMO system which allows
the WiMAX architecture to adapt the use of multiple
antenna techniques for example beam forming, space time
coding and spatial multiplexing
According to [8, 9] the MAC layer of WiMAX supports
not only fixed bit rates but also variable bit rates, real-time
and non-real-time traffic flows and also supports the best
effort data traffic.
The most innovative security features that are presently
used wireless access schemes are integrated in Mobile
WiMAX for voice, data and multimedia services. Such
features comprise, Advance Encryption Standard (AES)
based authentication and encryption, Cipher-based
Message Authentication Code (CMAC) and Hashed
Message Authentication Code (HMAC) based control
message protection schemes and Extensible
Authentication Protocol (EAP) based authentication. Thus
supporting varied user credentials such as digital
certificates, username and user password schemes, smart
cards and SIM/USIM cards [7].
The above features needs further investigation in terms
SNR on BER over various channels.
This paper is structured as follows. The introduction is
presented in chapters I. In chapter II the system model is
presented, while chapter III presents the methodology.
Section IV presents the simulation results and conclusion
is drawn in chapter V.
II. WIMAX OVERVIEW
The WiMAX Forum Network Group (WiMAX NWG)
developed the network architecture of WiMAX to
guarantee interoperability among diverse vendors and
their broadband equipment [8].
WiMAX (Worldwide Interoperability for Microwave
Access) is based on the IEEE 802.16 standard, formulated
to offer a shared basis for wireless connectivity in fixed,
portable, and mobile environment. WiMAX is a scalable
digital wireless access 4G technology intended to deliver
high throughput over long distances “wireless
metropolitan area networks” WMAN [5]. As a wireless
broadband technology, WiMAX perform similar to Wi-Fi
(IEEE 802.11) networks providing QoS (Quality of
Service) and coverage for fixed and mobile networks. It is
intended to be a complement or a competitor to cellular
technologies such as LTE and UMTS.
Interoperability is the key objective of WiMAX which
ensures that, equipment from different vendors
interoperates in the framework without difficulty.
International standards do not define boundaries for
subscriber module usage.
The systems is capable of covering large areas due to
various modulation schemes adopted such as QPSK, 16 –
QAM and 64 – QAM.
It is capable of providing enormous capacity for
bandwidth intensive applications compared to Universal
Mobile Telecommunication System (UMTS) and Global
System for Mobile communication (GSM).
User integrity, authentication and confidentiality are
assured through encryption and authentication protocols
such as AES adopted by the WiMAX standard. Numerous
system designs including Point-to-Point, Point-to-
Multipoint and Permeating coverage is permissible. It
provides communication for different traffic including
VoIP, multimedia applications and data, whiles ensuring a
higher degree of QoS.
Mobility is the key feature of the IEEE 802.16e and IEEE
802.16j standards due to Scalable Orthogonal Frequency
Division Multiple Access (SOFDMA) and Multiple Input
and Multiple Output (MIMO) techniques adopted at the
physical layer [5, 6, 8].
III. PAGE STYLE
The data bit stream of OFDM is divided into N data
streams using a rate of with each being parallel to each
other. i.e., the available spectrum must be divided into
several narrow sub-channels. Equalization becomes very
simple due to flat fading. Cyclic prefix, which is a copy of
last part of the OFDM symbol, is used in our model to
mitigate inter-symbol and inter-carrier interference while
maintaining orthogonality even though time-dispersive
channel is used for communication. This causes the
transmitted signal to appear periodic.
At a unique frequency, the individual streams are mapped
into tones and combined via IFFT (Inverse Fast Fourier
Transform) to produce the signal in time domain
A. OFDMA Transmitter
At the transmission stage, the signal in low pass is given
by:
 
0
1
2
0
1
( ) ( ) 1
N
j f t
k
k
s
x t s n e
T





With t defined within the interval;
( 1)
ss
nT t n T  
,
where
k
s
representing the baseband modulated signal.
The orthogonality of each subcarrier in OFDM according
to [15] is given by:

IJARCCE
ISSN (Online) 2278-1021
ISSN (Print) 2319 5940
International Journal of Advanced Research in Computer and Communication Engineering
Vol. 4, Issue 12, December 2015
Copyright to IJARCCE DOI 10.17148/IJARCCE.2015.41210 46
 
0
2 ( )
0
1 ,
1
2
0 ,
s
k
T
j f f t
s
kl
e dt
kl
T









With the minimum spacing given by:
 
1
1
3
kk
s
f f f
T

   
Hence by substitution, equation one becomes:
   
1
2
0
1
( ) 4
s
t
N
jk
T
k
k
s
x t s n e
T





OFDMA Transmitter
Finding the FT (Fourier Transform) of the signal
produces the frequency version the baseband signal given
by:
 
 
1
2
1
1
( , )
s
s
nT
j ft
s
nT
s
X f nT x t e dt
T




 
1
0
( )sin ( ) 5
n
N
j
ks
k
e s n c fT k






Sampling
 
1
,
s
X f nT
at
k
f
T

produces
 
k
sn
.
Thus:
   
, 6
sk
s
k
X nT s n
T




Where
0,1,2, 1kN
For the channel it is we assumed
 
;ht
to be the time
invariant response within the interval
0
cp
T  
limited
to the cyclic prefix length. The received signal
 
rt
is
given by:
          
0
( ) 7
cp
T
r t h x t h t x t d z t       

With
 
zt
representing additive complex and white
Gaussian noise.
B. OFDMA Receiver
For the receiver model, we assume a filter cache which is
matched to the last part of the waveform
 
k
t
given by:
 
 
 
, (0 )
8
0 , otherwise
k cp
k
T t t T T
t


    



OFDMA Receiver
By adding cyclic prefix, the transmitter utilizes rectangular
pulse modulated waveforms given by [12]:
 
 
 
2
1
, (0, )
9
0, otherwise
cp
W
j k t T
N
cp
k
e t T
TT
t










On a carrier frequency
W
k
N
, where
N
represent the
number of subcarriers,
W
represent bandwidth,
t
represent symbol length and
cp
T
represent the length of
cyclic prefix while
 
10
cp
N
TT
W

However,
 
 
11
kk
N
t
W
   
When
0
cp
tT
(i.e. when
t
is limited to the cyclic
prefix)
Therefore the baseband signal for OFDM symbol number
n
to be transmitted is given by:
     
1
,
0
12
N
n k n k
k
x t i t nT


  

With
0, 1, 2, 1,
, , , ,
n n n N n
i i i i


being a set of constellation
points complex numbers. For infinite sequence of OFDM
symbols
 
xt
to be transmitted,
     
1
,
0
( ) 13
N
n k n k
n n k
x t x t i t nT
  
  
   
  

IJARCCE
ISSN (Online) 2278-1021
ISSN (Print) 2319 5940
International Journal of Advanced Research in Computer and Communication Engineering
Vol. 4, Issue 12, December 2015
Copyright to IJARCCE DOI 10.17148/IJARCCE.2015.41210 47
IV. SIMULATION RESULTS
The simulation model comprises of a downlink PUSC
(partial usage of sub-channels) Physical Layer
communication between two MS (mobile stations) and a
BS (base station) based on the IEEE 802.16j standard. It
support all the mandatory modulation and coding schemes
offered by the standard, and demonstrate the variable size
capability of Embedded MATLAB, Simulink,
Communication System Toolbox and DSP System
Toolbox [16].
For simplicity, the model is limited to two MS while
applying FFT size of 1024. The paper specified 1024
subcarriers out of which 720 are reserved for user data
carriage and the remaining 304 kept for pilots and guards.
The 720 subcarriers are grouped into 30 sub-channels with
each channel holding 24 subcarriers. This allows for
efficient allocations of data carriers to different MS with a
sub-channel being the smallest unit permissible for
allocation.
The model allows for dynamic allocation of the sub-
channels or frequency resources to MS, thus, the BS is
able to dynamically adjust sub-channel allocation to MS1
and MS2 while the system is still running. For instance,
sub-channels 10 to 15 allocated to MS1 and sub-channels
16 to 28 allocated to MS2 in one burst can adjust to
becoming sub-channels 0 to 10 allocated to MS1 and sub-
channels 11 to 20 allocated to MS2 in another burst. This
initiates variable size signalling since data transmitted is
directly proportional to sub-channels allocated.
First, we perform our simulations by transmitting over a
non-fading pure AWGN (Additive White Gaussian Noise)
channel while adjusting the SNR at various levels. The
result of this is demonstrated in figure 4.1 and figure 4.2
for MS1 and MS2 respectively.
Figure 4.1 Effect of SNR on BER for various MCSs in
AWGN channel, (MS1).
The two figures showing the BER performance for MS1
and MS2 state clearly that, the performance improve as
SNR values goes high. There is no significant difference
between the performance of MS1 and MS2 in AWGN
channel without multipath fading. For both MSs, QPSK-
1/2 obtain optimum performance for SNR value at 8dB
and thereafter while 64 QAM-3/4 obtain best performance
at 24dB SNR value and subsequent values.
Figure 4.2 Effect of SNR on BER for various MCSs in
AWGN channel, (MS2)
Next, a Flat Fading Rician Channel is introduced in
AWGN environment while keeping all parameters
constant. Figures 4.3 and 4.4 show the BER performance
recorded for MS1 and MS2 respectively. It is observed
that, Flat Fading Rician channel has a significant effect on
the fidelity of the signal the signal received. Higher signal
power with respect to the noise factor is required for
optimum performance compared to when pure AWGN is
considered.
Figure 4.3 Effect of SNR on BER over a Flat Fading
Rician + AWGN Channel, (MS1)
Figure 4.4 Effect of SNR on BER over Flat Fading
Rician + AWGN Channel, (MS2)
For MS1, it is realized that, QPSK-1/2 records optimal
performance at 10 dB SNR and beyond while 64 QAM-
6 8 10 12 14 16 18 20 22 24 26 28 30 32
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
SNR ( dB)
BER
QPSK-1/2
QPSK-3/4
16 QAM-1/2
16 QAM-3/4
64 QAM-1/2
64 QAM-2/3
64 QAM-3/4
6 8 10 12 14 16 18 20 22 24 26 28 30 32
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
BER
SNR ( dB)
QPSK-1/2
QPSK-3/4
16 QAM-1/2
16 QAM-3/4
64 QAM-1/2
64 QAM-2/3
64 QAM-3/4
6 8 10 12 14 16 18 20 22 24 26 28 30 32
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
BER
SNR ( dB)
QPSK-1/2
QPSK-3/4
16 QAM-1/2
16 QAM-3/4
64 QAM-1/2
64 QAM-2/3
64 QAM-3/4
6 8 10 12 14 16 18 20 22 24 26 28 30 32
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
BER
SNR ( dB)
QPSK-1/2
QPSK-3/4
16 QAM-1/2
16 QAM-3/4
64 QAM-1/2
64 QAM-2/3
64 QAM-3/4

IJARCCE
ISSN (Online) 2278-1021
ISSN (Print) 2319 5940
International Journal of Advanced Research in Computer and Communication Engineering
Vol. 4, Issue 12, December 2015
Copyright to IJARCCE DOI 10.17148/IJARCCE.2015.41210 48
3/4 records its optimal performance at 27dB SNR and
beyond. On the hand, for MS2, QPSK recorded maximum
performance at 8dB SNR while 64 QAM recorded level
best performance at 25dB SNR which varied marginally
with that of AWGN channel.
Figures 4.5 and 4.6 depict the BER performance of
Frequency Selective Rician fading channel with AWGN.
For MS1, the performance of QPSK-1/2 (least modulation
and coding rate) reached level best at 16dB SNR value
while 64 QAM-3/4 (highest modulation and coding rate)
obtained optimal performance at 36dB SNR value and
thereafter. For MS2, QPSK-1/2 optimal performance at
18dB SNR values and 64 QAM-3/4 recorded most
favourable performance at as high as 52dB values of SNR
which does not reach zero BER at this point and beyond.
This clearly indicate that, Frequency Selective Rician
fading channels offer stiffer opposition to signal
transmission compared to Flat fading Rician and AWGN
channels.
Figure 4.5 Effect of SNR on BER over Frequency
Selective Rician + AWGN Channel, (MS1)
Figure 4.6 Effect of SNR on BER over Frequency
Selective Rician + AWGN Channel, (MS2)
BER performance of transmission over Rayleigh fading
with AWGN channel is analyzed next. From figure 4.7
which depicts the performance of MS1, QPSK-1/2 (least
modulation and coding rate) recorded optimal
performance at 8dB and higher values of SNR while 64
QAM-3/4 (highest modulation and coding rate) recorded
optimal performance with SNR values at 34dB and higher.
For MS2 depicted in figure 4.8, as QPSK-1/2 records
optimal performance at 8dB and higher values of SNR, 64
QAM-3/4 records best level performance which
approaches but never reaches zero BER at as high as 40dB
and higher SNR values. This indicates that, Rayleigh
Fading channel requires higher signal power relative to
noise for optimum performance compared to Rician and
AWGN channels.
Figure 4.7 Effect of SNR on BER over Rayleigh + AWGN
Channel, (MS1)
Figure 4.9 Effect of SNR on BER over Rayleigh + AWGN
Channel, (MS2)
After examining the BER performance, we investigate the
spectral efficiency against increase in SNR for MS1 over
the various channels under study. The average spectral
efficiency is seen as the number of successful bits
transferred per second per hertz of bandwidth [14]. For
each MCS, the spectral efficiency is plotted as a function
of SNR while determining the values of SNR that record
optimal spectral efficiency.
It is evident from figure 4.9 that, the maximum spectral
efficiency for AMC over AWGN channel is obtained as
8.789 bps/Hz at SNR values greater than or equal to 24 dB.
However, for individual MCSs depicted in figure 4.10, the
maximum spectral efficiency vary increasingly from
QPSK-1/2 recording 1.953 bps/Hz at 8 dB SNR value to
64 QAM-3/4 recording 8.789 bps/Hz at 24 dB SNR value.
An increase in code rate produced an increase in spectral
efficiency as the SNR requirement also goes high.
15 20 25 30 35 40 45 50 55
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
BER
SNR ( dB)
QPSK-1/2
QPSK-3/4
16 QAM-1/2
16 QAM-3/4
64 QAM-1/2
64 QAM-2/3
64 QAM-3/4
15 20 25 30 35 40 45 50 55
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
BER
SNR ( dB)
QPSK-1/2
QPSK-3/4
16 QAM-1/2
16 QAM-3/4
64 QAM-1/2
64 QAM-2/3
64 QAM-3/4
5 10 15 20 25 30 35 40 45
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
BER
SNR ( dB)
QPSK-1/2
QPSK-3/4
16 QAM-1/2
16 QAM-3/4
64 QAM-1/2
64 QAM-2/3
64 QAM-3/4
5 10 15 20 25 30 35 40 45
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
BER
SNR ( dB)
QPSK-1/2
QPSK-3/4
16 QAM-1/2
16 QAM-3/4
64 QAM-1/2
64 QAM-2/3
64 QAM-3/4