Analysis of Wavelet Based OFDM System
Analysis of Wavelet Based OFDM System
Analysis of Wavelet Based OFDM System
(1)
Where { | 0 1}
k
x k N a sequence in discrete time domain is { | 0 1} Xm m N are
complex numbers in discrete frequency domain that are obtained by applications of digital
modulation methods. The cyclic prefix is lastly added before transmission to minimize the
Inter Symbol Interference (ISI) and Inter Carrier Interference (ICI) [1], [2], [3].
At the receiver side, the process is reversed to obtain the decoded data. The cyclic prefix is
removed to obtain the data in the discrete time domain and then processed to FFT for data
MOD.
(QAM, PSK)
S/P
CONV.
IFFT
CYCLIC
PREFIX
ADDITION
N
CHANNEL
CYCLIC
PREFIX
REMOVAL
FFT
P/S
CONV. DEMOD.
4
recovery. The output of the FFT is the sum of the received signal in discrete frequency
domain as follows:
1
2 /
0
N
j km N
m k
k
X x e
(2)
After FFT, signal is converted back to parallel form and demodulated to yield the transmitted
signal back.
3. DWT BASED OFDM SYSTEM
Wavelet transform is a tool for studying signals in the joint time-frequency domains. So it is
capable of providing the time and frequency information simultaneously, hence giving a
time-frequency representation of the signal. The use of wavelet promises to reduce the ISI
and ICI without the usage of cyclic prefix as used in FFT Based OFDM [1], [3]. The
transform is implemented using filters. There are two filters pairs one for analysis pair also
called wavelet transform pair is a low pass filter while the other is a high pass filter. Figure 3
shows the block diagram for the implementation of wavelet transform upto one level only
therefore only one pair of filters (a high-pass filter and a low-pass filter) is used. Each sub
stream of data is sub-sampled by two as shown. The wavelet transforms divides the signal
into the approximate coefficients and detail coefficients as shown in figure 3.
Approximation coefficients
x[n] Detail coefficients
Figure 3: Block diagram of Filter Analysis [3]
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:
[ ] ( * )[ ] [ ] [ ]
k
y n x g n x k g n k
(3)
The signal is also decomposed simultaneously using a high pass filter h. The output gives
the detail coefficients (from the high-pass filter) and approximation coefficients (from the
low-pass). It is important that the two filters are related to each other and they are known as a
quadrature mirror filters [3]. 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:
[ ] ( * )[ ] [ ] [2 ]
low
k
y n x g n x k g n k
(4)
g[n]
h[n]
2
2
5
[ ] ( * )[ ] [ ] [2 ]
high
k
y n x h n x k h n k
(5)
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 frequency resolution has been doubled.
The transceiver of DWT based OFDM is shown in Figure 4. In a DWT based system,
inverse discrete wavelet transform (IDWT) and discrete wavelet transform (DWT)
replace the IFFT and FFT of FFT-OFDM system in modulation and demodulation
processes.
i/p o/p
Figure 4: DWT Based OFDM [3]
The output of the inverse discrete wavelet transform (IDWT) can be represented as:
/ 2
0 0
( ) 2 (2 )
n m m
m k
m n
d k D n
(6)
Where
n
m
D are the wavelet coefficients and (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 signals in scale and position or time. Xm
At the receiver side, the process is inversed. The output of discrete wavelet transform (DWT)
is
1
/ 2
0
( )2 (2 )
N
n m m
m k
k
D d k n
(7)
Therefore the QAM (Quadrature Amplitude Mirror) is a filter most commonly used to
implement filter bank that splits an input signal into two bands. Multirate signal processing is
used for the practical applications in signal processing to save costs, processing time and
many other practical reasons.
4. REVIEW ANALYSIS OF DIFFERENT OFDM SYSTEMS
1. The wavelet families Biorthogonal, Reverse-biorthogonal and Daubechies are
compared with FFT-OFDM and found that biorthogonal5.5 is superior among all. It
outperforms FFT and Daubechies by about 2dB and biorthogonal3.3 by 8 dB at 0.001
S
/
P
Constellation
Modulation
I
D
W
T
Channel
Constellation
Demodulation
D
W
T
P
/
S
6
Bit Error Rate. Using Reverse biorthogonal, the wavelet family reverse
biorthogonal3.3 has the least error as compared to others. At BER target of 0.001, it
outperforms FFT and Daubechies approximately 4 dB and reverse biorthogonal5.5
about 6 dB. Also the Daubechies families like db2 and db4 perform overlapping
curves with FFT-OFDM. This is possible since their orthogonality follows FFTs
orthogonality at small lter length such as 2 and 4 in the presence of additive white
Gaussian noise. Therefore DWT-OFDM is superior to FFT-OFDM when the system
use biorthogonal5.5 or reverse biorthogonal3.3 wavelet family [1].
Table 1: Analysis of FFT and DWT based OFDM
FFT-OFDM DWT-OFDM
Cyclic prefix Used Not used
Wavelet family - Biorthogonal, reverse biorthogonal
and daubechies
Bit error rate More Less
Orthonormal bases and
perfect reconstruction
properties
- Satisfied
2. The Haar and Daubechies based orthonormal wavelets are compared with
conventional OFDM and found that they are capable of reducing the power of
intersymbol interference (ISI) and intercarrier interference (ICI) due to the loss in
orthogonality between the carriers as a result of the multipath wireless channel. Also
analyzed that for the Wavelet Based MC transmission, as the number of carriers
increases ICI power is reduced [2]. Also orthonormal wavelets derived from Haar
QAMs provide good performance achievements in reducing both the ISI and ICI [2].
Table 2: Analysis of FFT and DWT based OFDM
Conventional
OFDM
Wavelet based OFDM
Wavelet family - Haar, daubechies
Average ISI reduced slightly reduced
Average ICI reduced Highly reduced
Bandwidth Divided uniformly Divided non-uniformly
Use of
bandwidth
- Large subchannel bandwidth used to
transmit high data rates i.e. video signals
and small subchannel bandwidth used to
transmit low data rates i.e. voice signal
3. For the FFT-OFDM system , AWGN, Rayleigh and Rician fading channels were
considered and analyzed that if the energy per bit to noise ratio was increases in
any system, the bit error rate was decreases accordingly. As the bit error rate was
less in wavelet based system it performs better compared to FFT based system
for any value of Eb/No [3].
7
Table 3: Analysis of FFT-OFDM and DWT-OFDM system
FFT Based OFDM DWT Based OFDM
Cyclic prefix With cyclic prefix Without cyclic prefix
Bit error rate more Less
Modulation methods BPSK, DBPSK BPSK, DBPSK
Channels AWGN, Rayleigh and
Rician fading channels
AWGN, Rayleigh and Rician
fading channels
4. Comparing with DWT-OFDM, WPT-OFDM and FFT-OFDM, DWT-OFDM gives
better performance than others because DWT-OFDM using zero-padding for detail
coefficients and also satisfies orthonormal bases and perfect reconstruction properties
[4]. The zero padding spreads the spectrum of the source symbols uniformly [9].
Table 4: Analysis of FFT-OFDM and DWT-OFDM system
FFT-OFDM WPT-OFDM DWT-OFDM
Cyclic prefix Used Not used Not used
Bit error rate More Less Less
Wavelet family - Biorthogonal, reverse
biorthogonal and
daubechies
Biorthogonal,
reverse
biorthogonal and
daubechies
5. Conclusion
This paper presents the analysis and review of OFDM techniques. The Discrete Wavelet
Transform is used instead of FFT which increases the orthogonality. There is no need for
using cyclic prefix because of excellent orthogonality offered by DWT, which in its order
reduces the system complexity, increases the transmission rate and increases spectral
efficiency.
After comparison between DWT based OFDM system with FFT based OFDM system, it was
found that DWT based OFDM system has better BER performance than that of the FFT
based OFDM for BPSK and DBPSK types of the modulations in the Additive White
Gaussian Noise channel, Rayleigh and Rician fading channels and as Eb/No ratio is
increased, bit error rate is decreases.
Also we analyzed that as the number of carriers increases, the Inter Carrier Interference
power is reduced dramatically for Wavelet Based multicarrier transmission system compare
to FFT-OFDM system. Therefore DWT based OFDM system is superior to the Fourier
transform based OFDM system especially when the system uses biorthogonal5.5 or reverse
biorthogonal3.3 wavelet family.
8
References
[1] Khaizuran Abdullah 1 and Zahir M. Hussain1, SMIEEE, Studies on DWT-OFDM and
FFT-OFDM Systems, SMIEEE, International Conference on Communication,
Computer and Power (ICCCP09) Muscat, February, 2009.
[2] B.G.Negash and H.Nikookar, Wavelet Based OFDM for Wireless Channels,
International Research Center for Telecommunication Transmission and Radar, Faculty
of Information Technology and Systems, Delft University of Technology, 2001.
[3] Swati Sharma1 and Sanjeev Kumar2, BER Performance Evaluation of FFT-OFDM
and DWT-OFDM, International Journal of Network and Mobile Technologies ISSN
2229-9114 Electronic Version VOL 2 / ISSUE 2 / MAY 2011.
[4] Khaizuran Abdullah, Amin Z. Sadik, and Zahir M. Hussain, SMIEEE, On the DWT-
and WPT- OFDM versus FFT-OFDM, School of Electrical & Computer Engineering,
RMIT University, Melbourne, Australia, 2009.
[5] Haitham J. Taha and M. F. M. Salleh, Multi-carrier Transmission Techniques for
Wireless Communication Systems: A Survey, WSEAS Transactions on
Communications, Issue 5, Volume 8, May 2009.
[6] Upena Dalal, Wireless Communication, Oxford University press, july 2009, pp.365
408.
[7] M. Weeks, Digital Signal Processing Using Matlab and Wavelets, Innity Science
Press LLC, 2007, pp.275- 336.
[8] Oppenheim, Schafer with Buck, Discrete-Time Signal Processing, PEARSON
Education, 2
nd
edition, 2005.
[9] Jonathan H. Manton, An OFDM interpretation of zero padded block transmissions,
ARC Special Research Centre for Ultra-Broadband Information Networks, Department
of Electrical and Electronic Engineering, The University of Melbourne, Melbourne,
Victoria 3010, Australia, 2002.