International Conference on Computational Intelligence and Multimedia Applications 2007

A Comparative Performance Analysis of OFDM using MATLAB

Simulation with M-PSK and M-QAM Mapping

Jigisha N. Patel Prof.(Mrs.) Upena D.Dalal

Electronics Engg. Dept. Electronics Engg. Dept.
SVNIT, Surat-395007 SVNIT, Surat-395007
e-mail: jigisha_js@yahoo.co.in e-mail: upena_dalal@yahoo.com

Abstract
In wireless communication, concept of parallel transmission of symbols is applied to
achieve high throughput and better transmission quality. Orthogonal Frequency Division
Multiplexing (OFDM) is one of the techniques for parallel transmission. The idea of OFDM
is to split the total transmission bandwidth into a number of orthogonal subcarriers in order
to transmit the symbols using these subcarriers in parallel. In this paper, proposed OFDM
system design is simulated using MATLAB simulink toolbox. The digital modulation schemes
such as M-PSK (M-ary Phase Shift Keying) and M-QAM (M-ary Quadrature Amplitude
Modulation), which provide way of parallel transmission, are compared to analyze the BER
performance of designed OFDM system. Mentioned schemes used in OFDM system can be
selected on the basis of the requirement of power or spectrum efficiency and BER analysis.

1. Introduction
Many methods are proposed to combat the multipath effects in wireless
communication. One of the solutions to combat Inter Symbol Interference (ISI) is
multicarrier modulation for data transmission [1], [3], [11], that is Orthogonal
Frequency Division Multiplexing (OFDM). The analysis of Bit Error Rate (BER)
performance suggests, OFDM is better than Code Division Multiple Access (CDMA)
which is mostly incorporated in existing 3G systems [3], [11].
The aim of OFDM is to divide the wide frequency selectivity of fading channels into
multiple flat fading channels [1], [11]. The idea of using a Discrete Fourier Transform
(DFT) for the generation and reception of OFDM signals eliminates the requirement of
banks of analog sub carrier oscillators [5] [10]. Orthogonality property allows multiple
information signals to be transmitted in parallel over a common channel and detected,
without interference. In OFDM spectrum each subchannel has a peak at the subcarrier
frequency and nulls evenly spaced with a frequency gap equal to the carrier spacing ∆f
= 1/T s, where T s is OFDM symbol duration [1], [2], [5]. Another characteristic of
orthogonality is that each carrier has an integer number of sine wave cycles in one bit
period [8].
Although OFDM enables simple equalization, it is sensitive to carrier frequency offset [2],
[7]. The peak to average ratio (PAR) of the transmitted signal power is large [4], [5], [7], [8].
OFDM system performance can be improved by channel coding [1], [6].

2. Model Design of OFDM Transceiver Using MATLAB/SIMULINK

The OFDM system is modeled using MATLAB/SIMULUNK to allow various parameters
of the system to be varied and tested. The following OFDM system parameters are
considered for the simulation.

0-7695-3050-8/07 $25.00 © 2007 IEEE 406

DOI 10.1109/ICCIMA.2007.142
 Bit rate R = 1/T : 1 Mbps
Data mapping : M-PSK and M-QAM
IFFT, FFT size : 64-point
Channel used : AWGN
Guard Interval size : IFFT size/4 = 16 samples
OFDM transmitted frame size: 64+16 = 80
The system model for OFDM with M-PSK mapping is shown in Figure 1, representing the
following blocks. M-PSK block can be replaced by M-QAM block for further comparison.

Binary source: The random Bernoulli binary generator generates binary data that is frame
based. In data output, 48 samples per frame are used, and data rate is 1 Mbps.
Data mapping: The input data stream is available serially, converted into parallel stream
according to digital modulation scheme. The data is transmitted in parallel by assigning each
data word to one carrier in the transmission. Once each subcarrier has been allocated
symbols, they are phase mapped according to modulation scheme, which is then represented
by a complex In-phase and Quadrature-phase (I-Q) vector.
The constellation diagrams of different M-PSK and M-QAM mapping are shown in
Figure 2. Consider QPSK mapping in M-PSK block of proposed model, which maps 2 bits
per symbol into phase, as shown in Figure 2(a). Each combination of 2 bits of data
corresponds to a unique I-Q vector. In M-PSK block, by changing bits per symbol, we can
map the data for 8-PSK, 16-PSK etc. By moving to higher order constellation, it is possible
to transmit more bits per symbol in parallel resulting in high speed communication. The use
of phase shift keying produces constant amplitude signal and reduce problems with
amplitude fluctuation due to fading.
M-QAM modulation can be considered as combination of ASK (Amplitude Shift Keying)
and M-PSK. Digital M-PSK is a special case of M-QAM, where the amplitude of the
modulated signal is constant. In M-QAM, constellation points are usually arranged in a
square grid with equal horizontal and vertical spacing as shown in Figure 2(c) and Figure
2(d), although other configurations are also possible [9]. If data rates beyond those offered by
8-PSK are required, it is more usual to move to M-QAM since it achieves a greater distance
between adjacent points in the I-Q plane by distributing the points more evenly. In M-QAM
the location of constellation points no longer indicate the same amplitude and so the
demodulator must now correctly detect phase and amplitude, rather than just phase.

Figure 1. Block diagram of OFDM transceiver

407
 In general, the selection of modulation scheme applied to each subchannel depends solely
on the compromise between the data rate requirement and transmission robustness.

IFFT-Frequency domain to time domain conversion: The IFFT converts frequency

domain data into time domain signal and at the same time maintains the orthogonality of
subcarriers. The real signal output can be generated by arranging conjugate subcarriers [4] as
shown in Figure 3(b).
In this stage, IFFT mapping, zero pad, and selector blocks are included. Zero pad block
adds zeros to adjust the IFFT bin size of length L, as the number of subcarriers may be less
than bin size. Selector block reorders the subcarriers. The IFFT bin setting, for complex
OFDM signal for the given design, is shown in Figure 3(a). The IFFT block computes the
Inverse Fast Fourier Transform (IFFT) of length L points, where L must be a power of 2 [8].
Guard period: The effect of ISI on an OFDM signal can be eliminated by the addition of
a guard period at the start of each symbol [5]. This guard period is a cyclic copy that extends
the length of the symbol waveform. The guard period adds time overhead, decreasing the
overall spectral efficiency of the system. Guard duration should be longer than channel delay
spread [5]. After the guard band has been added, the symbols are converted into serial form.
One frame length duration T = Ts + Tg , where Ts = NT, N = number of carriers. This is the
OFDM base band signal, which can be up converted to required transmission frequency.
An AWGN channel model is then applied to transmitted signal. The model allows for the
Signal to Noise Ratio (SNR) variation. The receiver performs the reverse operation of the
transmitter. The receiver consists of removal of guard band, FFT, removal of zero padding
and demapping of data.

3. Simulation result
The performance of a data transmission system is usually analyzed and measured in terms
of the probability of error at given bit rate and SNR. The parameter Eb/No, where Eb is bit
energy and No is noise energy, is adjusted every time by changing noise in the designed
channel.

(a) 4-PSK (b) 8-PSK (c) 8-QAM (d) 16 QAM

Figure 2. Constellation diagrams 4/8-PSK and 8/16-QAM

(a) Complex Output OFDM Signal (b) Real Output OFDM Signal

Figure 3. Concept of IFFT bin setting used in the simulation

408
 For particular Eb/No value, system is simulated and corresponding probability of error is
noted. The proposed design is simulated with necessary parameter changes for QPSK, 8-PSK
and 16-PSK. As shown in Figure 4, if we go on increasing the Eb/No value, BER reduces. In
comparison of BER performance for M-PSK, it is observed that use of a higher M-ary
constellation is better for high capacity transmission but the drawback is that the points on
constellation are closer which makes the transmission less robust to errors with same SNR.
For OFDM with QPSK simulation, constellation diagram of transmitted signal and
received signal is shown in Figure 5. The OFDM with 8-QAM and 16-QAM mapping
simulation are analyzed for BER performance and compared with 8-PSK and 16-PSK
systems simulation as shown in Figure 6.

0.5
4-PSK
8-PSK
0.4
16-PSK

0.3
BER

0.2

0.1

0
0 10 20 30 40
Eb/No (db)

Figure 4. BER performance comparison of 4/8/16-PSK

(a) Transmitted signal with 4 phases (b) Received signal with phase distortions

Figure 5. Constellation diagrams for QPSK mapped simulation

Figure 6. BER performance comparisons of 8/16-PSK and 8/16-QAM

409
4. Conclusion
OFDM is a powerful modulation technique used for high data rate, and is able to eliminate
ISI. It is computationally efficient due to the use of FFT techniques to implement modulation
and demodulation functions. The performance of OFDM is tested for two digital modulation
techniques namely M-PSK and M-QAM using MATLAB/SIMULINK toolbox. It is
observed from M-PSK BER plot that BER is less in case of 4-PSK for low Eb/No as
compared to 8- PSK and 16-PSK. Hence, high value of M-ary increases spectrum efficiency,
but easily affected by noise. So OFDM system with QPSK scheme is suitable for low
capacity, short distance application. While the OFDM with higher M ary modulation scheme
is used for large capacity, long distance application at the cost of slight increase in Eb/No.
The comparison of M-PSK and M-QAM indicates that, BER is large in M-PSK as
compared to M-QAM and it generally depends on applications. For higher value of M that is
for M > 16, QAM modulation scheme is used in OFDM. Similarly, results can be tested with
addition of channel coding block in the model design.

5. References
[1] John A.C Bingham, “Multicarrier modulation for data transmission: An idea whose time has come,” IEEE
Communication Magazine, May 1990, vol.28, issue 5, pp.5-14.
[2] Flemming Bjerge Frederiksen and Ramjee Prasad, “An Overview of OFDM and related techniques towards
development of future wireless multimedia communications,” Radio and Wireless Conference, 11-14 Aug
2002, IEEE, pp.19-22.

[3] Mehul Jain and M. Mani Roja, “Comparison of OFDM with CDMA System in Wireless Telecommunication
for multipath delay spread,” The first IEEE and IFIP International Conference in Central Asia, 26-29 Sept.
2005, pages 5.

[4] Burton R. Saltzberg, “Comparison of single carrier and multitone digital modulation for ADSL application,”
IEEE communication Magazine, Nov. 1998, vol.36, issue 11, pp. 114-121.

[5] William Y. Zou Yiyan Wu, “COFDM: An Overview,” IEEE Transaction on Broadcasting, March 1995,
vol.41, issue 1, pp. 1- 8.

[6] W.A.C. Fernando, R.M.A.P. Rajtheva, K.M.Ahmed, “Performance of coded OFDM with higher modulation
schemes,” Communication Technology Proceedings, Beijing, China, October 22-24, 1998, vol.2 pp. 2.

[7] Zhengdao Wang, “OFDM or single carrier block transmission,” IEEE Transaction on Communication,
March 2004, vol.52, issue 3, pp. 480-394.

[8] Wi-LAN, Wireless Data communications, “Wideband Orthogonal Frequency Divison Multiplexing (W-
OFDM),” White paper, Year 2000, version 1.0.

[9] J.A.Sills, “Maximum–Likelihood Modulation Classification for PSK/QAM,” Military Communications

Conference Proceedings, IEEE, Oct.-Nov. 1999, vol.1, pp.217-220.

[10] M. Schiibinger, S.R.Meier, “DSP-based signal processing for OFDM transmission,” Acoustics, Speech, and
Signal Processing Proceedings (ICASSP’01), IEEE, 7-11 May 2001, vol.2, pp.1249-1252.

[11] Bernard Sklar, Digital Communication Fundamentals and Applications, Second Edition, Pearson Education
Asia, 2000.

410