JOURNAL OF TELECOMMUNICATIONS, VOLUME 18, ISSUE 2, FEBRUARY 2013

Wavelet and FFT based BER Analysis for

Multicarrier Communication
*Shivaji Sinha and **Chakresh Kumar
AbstractIn 4G wireless communication systems, bandwidth is a precious parameter, and service providers are continuously
met with the challenge of accommodating more users with in a limited allocated bandwidth. To increase data rate of wireless
medium with higher performance, Recently DWT- OFDM is adopted in place of FFT OFDM for frequency translation in OFDM
system. Modulation schemes such as 4-PSK, DPSK & QAM have been used in the developed OFDM system for both DWT
and FFT based model. In this paper comparative studies on DWT-OFDM & FFT-OFDM systems are presented. Results also
show that DWT-OFDM is superior as compared to FFT-OFDM with regards to the BER performance, especially when it uses
bior 5.5 or rbior 3.3 wavelet family for AWGN channel. This shows that the DWT families better over the IFFT-FFT
implementation. The results presented in the paper are based on Matlab based simulations; a highly efficient tool for different
applications.
Index TermsDiscrete Wavelet Transform (DWT), Fast Fourier Transform (FFT), OFDM, AWGN, BER, SNR, Cyclic Prefix
(CP), DPSK, PSK, QAM.

1 INTRODUCTION
HE fundamental principle of the multicarrier (MCM
OFDM) modulation system is to decompose the high
rate data stream (bandwidth W) into N parallel lower
rate data streams or channels, one for each subcarrier.
Each sub-carrier is modulated with a classical modulation
scheme (such as Quadrature amplitude modulation or
Phase-shift keying) at a low symbol rate, maintaining
total data rates similar to conventional single-carrier
modulation schemes in the same bandwidth. A large
value of N makes the individual bandwidth (W/N) of
subcarriers narrower than the coherence bandwidth of
the channel [1].The main advantage of OFDM over single
carrier schemes is its ability to cope with multipath
channel conditions (for example, attenuation of high
frequencies in a long copper wire, narrowband interfe-
rence and frequency-selective fading due to multipath)
without use of complex equalization filters. OFDM may
be viewed as using many slowly modulated narrowband
signals rather than one rapidly-modulated wideband
signal. The choice of individual subcarriers is such that
they are orthogonal to each other, which allows for the
overlapping of subcarriers because the orthogonality en-
sures the separation of subcarriers at the Multiple Access,
where no spectral overlap of carriers is allowed. The or-
thogonality allows for efficient modulator and demodu-
lator implementation using the FFT algorithm on the re-
ceiver side, and inverse FFT on the sender side [2]. The
major contribution to the OFDM implementation was the
application of the Fast Fourier Transform (FFT) to the
modulation and demodulation processes.
Fortunately, the availability of digital signal processing
techniques makes it easy to implement it. Recently the
Wavelet Transform has also been proposed as a possible
transform to generate the sub channels in a multicarrier
system, with the advantage of flexibility of the transform
and the higher suppression of side lobes compared to the
side lobes of the rectangular window in the Fourier
Transform.[3],[5].

2 FOURIER TRANSFORM BASED OFDM SYSTEM
A typical block diagram of an OFDM system is shown in
fig. 1.









Fig. 1. Typical model of an OFDM transceivers with inverse and
forward transforms that can be substituted as FFT or DWT OFDM
The inverse and forward transform blocks in fig 1 can be
FFT, DCT or DWT based blocks.
This FFT-based sysem model for OFDM will not be dis-
cussed in detail as it is well known in the literature. Thus,
a brief description about it is presented here. The data d
k

first being processed by a constellation mapping. M-ary
QAM modulator is used for this work to map the raw
binary data to appropriate QAM symbols. These symbols
are then input into the IFFT block. This involves taking N
parallel streams of QAM symbols (N being the number of
sub-carriers used in the transmission of the data) and
T

- * Assistant Professor,Department of Electronics Engineering at J.S.S.
Academy of Technical Education Noida, India

- **Assistant Professor,Department of Electronics and communication at
Tezpur (central) University ,Tezpur, Assam,India



5


performing an IFFT operation on this parallel stream. The
output in discrete time domain is given as follows

.
(1)

where
( ) k n
X | 0 n N - 1 is a sequence in the discrete
time domain and
( ) m i
X | 0 i N 1 are complex
numbers in the discrete frequency domain. The cyclic
prefix (CP) is lastly added before transmission to
minimize the inter-symbol interference (ISI) and Inter
carrier Interference (ICI). If the channel impulse response
has a known length L, then the prefix consists simply of
copying the last L-1 values from each symbol and ap-
pending them in the same order to the front of the sym-
bol.
At the receiver, the process is reversed to obtain
the decoded data. The CP is removed to obtain the data in
the discrete time domain and then processed to FFT for
data recovery. The output of the FFT in the frequency
domain is given as follows



(2)

3 WAVELET
TRANSFORM BASED OFDM SYSTEM
As mentioned in the previous section, the inverse and
forward block transforms are flexible and can be substi-
tuted with FFT or DWT- OFDM block. In this section we
will describe wavelet based OFDM system.

Wavelets are wave forms of effectively limited dura-
tion that has an average value of zero and localization
both in time and frequency. They also possess the prop-
erty of orthogonality across scale and translation. The
comparative difference between wavelets and sine waves,
which are the basis of Fourier analysis is that sinusoids do
not have limited duration, they extend from minus to
plus infinity and where sinusoids are smooth and pre-
dictable, wavelets tend to be irregular and asymmetric.
As the well known technique of signal analysis Fourier
analysis consists of breaking up a signal into sine waves
of various frequencies, similarly, wavelet analysis is the
breaking up of a signal into shifted and scaled versions of
the original (or mother) wavelet. The Discrete Wavelet
Transform (DWT) is used in a variety of signal processing
applications, such as video compression, Internet com-
munications compression, object recognition and numeri-
cal analysis [4].


3.1. Discrete Wavelet Transform (DWT)
The DWT based transceiver for OFDM is shown in
fig. 2a. and fig. 2b.The transmitter first uses a digital
modulator (16-QAM) which maps the serial bits into
symbols converting d
k
into X
m
, within N parallel data X
m(i)

stream,where
( ) m i
X | 0 i N 1 as explained in fig 1.
The main task of the transmitter is to perform the discrete


Fig.2a. DWT-OFDM Transmitter Model

wavelet based modulation by constructing orthonormal
wavelets.
Each X
mi
is first converted to serial representation having
a vector XX which will next be transposed into CA. This
mean that CA not only its imaginary part have
inverting signs but also its form is changed to a parallel
matrix. Then, the signal is up-sampled and filtered by the
LPF coefficients or namely as approximated coefficients.
Since our aim is to have low frequency signals, the modu-
lated signals XX perform circular convolution with LPF
filter whereas the HPF filter also perform the convolution
with zeroes padding signals CD respectively. Note that
the HPF filter contains detailed coefficients or wavelet
coefficients. Different wavelet families have different fil-
ter length and values of approximated and detailed coef-
ficients. Both of these filters have to satisfy orthonormal
bases in order to operate as wavelet transform. In the
transmitter part, this signal is simulated using MATLAB
command
[X
k
] = idwt(CA,CD; wv)

where wv is the type of wavelet family.

The block diagram of DWT based 16-QAM de-
modulated receiver for OFDM system is shown in
fig. 2(b) to recover back the original data.











Fig.2b. DWT-OFDM Receiver Model

The reverse process for DWT is simulated in the receiver
using matlab command
[ca, cd] = dwt(Uk,, wv)
1
( 2 )
( ) ( )
0
1
n N
j i
N
m i k n
n
U U e
N
t

=
=

1
( 2 )
( ) ( )
0
1
n N
j i
N
k n m i
i
X X e
N
t

=
=

6



3.1 Filter Analysis
The advantage of wavelet transform over other tranforms
such as Fourier transform is that it is discrete both in time
as well as scale. The transform is implemented using
filters. One filter of the analysis (wavelet transform) pair
is a low-pass filter (LPF), while the other is a high-pass
filter (HPF) as show in Fig 3. Each filter has a down-
sampler after it, to make the transform efficient.


Fig. 3 Two channel filter bank showing perfect reconstruction
property

The DWT of a signal x is calculated by passing it through
a series of filters. First the samples are passed through a
low pass filter with impulse response g resulting in a
convolution of the two as given by the equation (3)




The signal is also decomposed simultaneously using a
high-pass filter h. The output gives the detail coeffi-
cients (from the high-pass filter) and approximation coef-
ficients (from the low-pass). It is important that the two
filters are related to eachother and they are known as a
quadrature mirror filters. However, since half the
frequencies of the signal have now been removed, half the
samples can be discarded according to Nyquists rule.
The filter outputs are then sub sampled by two. The
outputs of the low-pass filter and the high-pass filter are
shown below and are the convolutions of the input data
with the respective filter responses is given by the equa-
tion (4)








The decomposition has halved the time resolution since
only half of each filter output characterizes the signal.
However,each output has half the frequency band of the
input so the requency resolution has been doubled.

Fig. 3 also shows the implementation of wavelet trans-
form up to one level only and hence only one pair of fil-
ters (a high-pass filter and a low-pass filter) is used. Each
sub stream of data is sub-sampled by two.

One of the advantage of using wavelet transform is that
due to the overlapping nature of wavelet properties, the
wavelet based OFDM does not need cyclic prefix to deal
with delay spreads of the channel. As a result, it has
higher spectral containment than that of Fourier-based
OFDM [6]. The input data is processed as per FFT-OFDM.
However, the difference is that the system does not re-
quire CP to be added to the OFDM symbol. The output of
the inverse discrete wavelet transform (IDWT) can be
represented as:


(6)

Where (t) is the
wavelet function with compressed factor m times and
shifted n times for each subcarrier (number k, 0 k N
1). The wavelet coefficients are the representation of sig-
nals in scale and position or time. At the receiver side, the
process is inversed. The output of discrete wavelet trans-
form (DWT) is given by





Multiresolution analysis of wavelet theory allows to
represents wavelet and scaling functions by high and low
pass filters (HPF and LPF), respectively with impulse res-
ponses h[m] and g[m]. Therefore the wavelet transforma-
tion can be easily implemented using discrete time filters
[7][8].
4. SOFTWARE SIMULATION PARAMETERS

The performance of FFT-OFDM and DWT-OFDM has
been investigated by means of computer simulation
Parameters for the simulation are shown in table 1.

TABLE 1: SIMULATION PARAMETER



These parameters are loosely based on wireless Local
Area Network (WLAN) standard IEEE 802.11a, which is
used in 2.5,3.6 and 5 GHz frequency range.
5 RESULT ANALYSIS
Matlab simulated graphical results show the bit error rate
PARAMETER VALUE
Number of subcarrier 48
Number of pilot carrier 4
Number of total subcarrier 52
FFT length 64
Frame Size 1000 bits/frame
Modulation schemes QAM,PSK and DPSK
Channel encoding Convolution
Wavelet family used Bior 5.5, rbio 3.3
[ ] ( )[ ] ( ) ( )
k
k
y n x g n x k g n k
=+
=
= - =

[ ] ( )[ ] ( ) (2 )
k
low
k
y n x g n x k g n k
=+
=
= - =

/ 2
0 0
( ) 2 (2 )
n m m
m k
m n
d k D n

= =
=

[ ] ( )[ ] ( ) ( )
k
high
k
y n x h n x k h n k
=+
=
= - =

1
/ 2
0
( )2 (2 )
n
n m m
m k
k
D d k n

=
=

7
(4)
(5)
(7)
(3)


probabilities for different values of Sgnal to ratios for
both the DWT and FFT based OFDM sytems. In the sys-
tem implemented convolution coding is used for error
correction, pilot carriers have also been used for timing
synchronization. Different wavelets are available for the
implementation of discrete wavelet transform and we
have considered bior 5.5 and rbior 3.3 wavelet family for
the system. The channel considered here is AWGN. The
results presented show the BER performance as a func-
tion of energy per bit to noise ratio.

In the implemented system, convolution coding is
used for error correction, pilot carriers have also been
used for timing synchronization. Different wavelets are
available for the implementation of discrete wavelet
transform and we have considered bior 5.5 and rbior 3.3
wavelet for the system. The channel considered here is
AWGN. The results presented show the BER performance
as a function of energy per bit to noise ratio.


Fig, 4. BER Performance of DWT-OFDM and FFT-OFDM with
4-PSK using bior 5.5 wavelet

Fig. 5a. BER Performance of DWT-OFDM and FFT-OFDM with
DPSK using bior 5.5 wavelet




Fig. 5b. BER Performance of DWT-OFDM and FFT-OFDM
with DPSK using rbio 3.3 wavelet


Fig. 6a. BER Performance of DWT-OFDM and FFT-OFDM with
4-QAM using bior 5.5 wavelet

Fig. 6b. BER Performance of DWT-OFDM and FFT-OFDM with
4-QAM using rbior 3.3


0 2 4 6 8 10 12 14
10
-4
10
-3
10
-2
10
-1
10
0
BER vs SNR with 4-PSK and bior5.5 wavelet
B
E
R
SNR (dB)

DWT-OFDM
FFT-OFDM
0 5 10 15 20 25 30
10
-4
10
-3
10
-2
10
-1
10
0
B
E
R
SNR (dB)
BER vs SNR with DPSK and bior5.5


DWT-OFDM
FFT-OFDM
0 5 10 15 20 25 30
10
-3
10
-2
10
-1
10
0
BER vs SNR with DPSK and rbio3.3 wavelet
B
E
R
SNR (dB)

DWT-OFDM
FFT-OFDM
0 1 2 3 4 5 6 7 8 9 10
10
-4
10
-3
10
-2
10
-1
10
0
BER vs SNR with 4-QAM and bior5.5 wavelet
B
E
R
SNR (dB)

DWT-OFDM
FFT-OFDM
0 1 2 3 4 5 6 7 8 9 10
10
-4
10
-3
10
-2
10
-1
10
0
BER vs SNR with 4-QAM and rbior3.3 wavelet
B
E
R
SNR (dB)

DWT-OFDM
FFT-OFDM
8




Fig. 7a. BER Performance of DWT-OFDM and FFT-OFDM
with 16-QAM using bior 5.5 wavelet


Fig. 7b. BER Performance of DWT-OFDM and FFT-OFDM
with 16-QAM using rbio 3.3 wavelet

As shown in figures 4,5,6 and 7 it is found that as the
energy per bit to noise ratio increases in any modulation
system a ireduction in bit error rate is encountered. Also
wavelet based system performs better in terms of bit error
rate probability by showing less bit error rate as
compared to FFT based system for any value of E
b
/N
o
in
all modulation schemes. This is so because the channel
impulse response length is long due to the insertion of
cyclic prefix in case of FFT while in case of DWT no need
of cyclic prefix is required so impulse response length is
small and hence Inter symbol interference(ISI) & Inter
carrier interference (ICI) is reduced to great extent. The
spectral containment of the channel is better and the data
throughput and power efficiency is also increased in
DWT based OFDM system because of not using cyclic
Prefix. The results shown in all the curves for different
modulation schemes could have been better if more no of
samples for the symbols were used. However this yields
longer simulation time. From the figures, it is seen that
biorthogonal families shows better BER results than FFT
based OFDM System. It is shown that bior 5.5 is not only
superior among other wavelet family in DWT based
OFDM but it also better perform than FFT based OFDM
system for the same higher SNR values. From fig. 7a it is
obsedrved that for the value of BER 10
-3
, bior 5.5 wavelet
system outperforms FFT by about 4 dB.

From the fig 6a,6b,7a and 7b, we observed that as the
modulation schemes is changed from PSK to QAM, the
BER also increases although it is better for DWT based
system as compared to FFT based system.


6 CONCLUSION
This paper compares the performance of Fourier transform
based and Wavelet based OFDM systems in terms of bit
error rate probability for different modulation schemes and
wavelets. From the performed simulations in the Additive
White Gaussian Noise channel, it was found that the DWT
based OFDM system has better performance than that of the
FFT-OFDM for all 4 types of the modulations used viz.
16-QAM, 4-QAM, DPSK and 4-PSK. It was also found that
DWT-OFDM outperformed FFT-OFDM for both type of
wavelet used. Wavelet based system was found having small
bit error rate probability than that of the Fourier transform
based system.

The purpose of the paper was to implement and find the
transform that performs better for different wavelet based
modulation schemes.
7 FUTURE WORK
The Future work may include implementation of this
system for different multipath channels using MIMO
OFDM system and also the modulation schemes. The
performance of the DWT system can also be checked for
various other wavelet families available.
8 REFERENCES
[1] H. W. Newlin, "Developments in the use of wavelets in
communication systems," Proc. of 1998 IEEE Military Commun.
Conf. (MILCOM '98), vol. 1, Bedford, Massachusetts, USA,
October 1998, pp. 343-349.
[2] Swati Sharma and Sanjeev Kumar, BER Performance Evalua-
tion of FFT-OFDM and DWT-OFDM, International Journal of
Network and Mobile Technologies, VOL2 / ISSUE2 / MAY2011.
[3] S. B. Weinstein, and P. M. Ebert, "Data transmission by fre-
quency division multiplexing using the discrete Fourier trans-
form", IEEE Transactions on Communication Technology, vol.
19, no. 5, pp. 628-634, 1971.
[4] Michael Weeks, Digital Signal Processing Using MATLAB and
Wavelets. Infinity Science Press LLC, 2007
0 2 4 6 8 10 12 14 16 18
10
-4
10
-3
10
-2
10
-1
10
0
BER VS SNR with 16-QAM and bior5.5 wavelet
B
E
R
SNR (dB)

DWT - OFDM
FFT - OFDM
0 2 4 6 8 10 12 14 16 18
10
-4
10
-3
10
-2
10
-1
10
0
BER vs SNR with 16-QAM and rbio3.3 wavelet
B
E
R
SNR (dB)

DWT-OFDM
FFT-OFDM
9


[5] A.Akansu, P.Duhamel, X.Lin, and M.de Courville, Orthogonal
Tran multiplexers in communication: a review, IEEE Transac-
tions on signal processing, vol.46no.4, pp.979-995, April 1998.
[6] K. Abdullah and Z.M. Hussain, "Studies on DWT-OFDM and
FFT-OFDM Systems" International Conference On Communica-
tion, Computer And Power (Icccp09) Muscat, February 15-18,
2009,SQU-2009 ISSN: 1813-419X.
[7] S. Baig, F. U. Rehman, and M. J. Mughal, Performance compari-
son of DFT, discrete wavelet packet and wavelet transforms, in
an OFDM transceiver for multipath fading channel, in Proceed-
ings of 9th International Multitopic Conference, INMIC05,
2005, pp. 1-6.


Shivaji Sinha received his Bechelor of Engineering in Electronics
and Communication Engineering in 2003 and Mas-
ter of Technology degree in VLSI design in 2011
respectively. For 10 years he is working in de-
partment of Electronics Engineering at J.S.S.
Academy of Technical Education, Noida, India
from 2003 to till date. He is presently working as
an Assistant professor. He is a Senior member of
Institution of Electronics & Telecommunication
engineers, New Delhi, Indian Society of Technical Education, New
Delhi, Institution of Communication Engineers & Information Tech-
nologies, New Delhi, The Institution of Engineers (IE), Kolka-
ta,India, International Association of Computer Science and Informa-
tion Technology,Singapore ,International Association of Engi-
neers(IAENG) and Computer Science Teachers Association(CSTA),
USA.His current research area is Mobile and Wireless Communica-
tion.

Chakresh Kumar received his Bechelor of Technology degree in
Electronics and Communication Engineering in 2006 and Master of
Technology degree from I.S.M Dhanbad in 2009
respectively. His specialization is in Optical
Communication. Presently he is working as an
Assistant Professor in the department of Elec-
tronics and communication Engineering at Tez-
pur (Central) University, Tezpur Assam, India.He
is member of no of National & Internation tech-
nical Societies.


10