08-08-2014, 02:10 PM
Design and Simulation of Wavelet OFDM with Wavelet Denoising on AWGN Channel
[attachment=66751]
Abstract
OFDM technique is widely used due to its high spectrum efficiency. Fast Fourier Transforms (FFT) has
been used to generate the orthogonal sub-carriers. Implementation complexity of FFT based OFDM, has led to replace
conventional FFT based OFDM by wavelet based OFDM. The main objective of this paper is to replace the
conventional FFT based OFDM system with some orthonormal wavelets. Comparing with the conventional OFDM, it
is found that the Haar and Daubechies-based orthonormal wavelets are capable of reconstructing the transmitted
symbol at the receiver side and effect of noise is reduced using wavelet denoising for different SNRs on AWGN
channel. In this paper, results are tabulated for transmitted symbols and reconstructed symbols after wavelet
denoising& compared for an analysis. The scatterplot analysis of DWT-OFDM is also presented for different SNRs.
INTRODUCTION
OFDM system is one of the most promising technologies
for current and future 4G wireless communications [1]. It
is a multi-carrier transmission technique [2], which divides
the available spectrum into many carriers, each one being
modulated by a low rate data stream. Conventional OFDM
system used IFFT and FFT to multiplex the signals in
parallel with reduced complexity algorithm at the
transmitter and receiver respectively. Many researchers
have investigated the use of wavelet based to replace
Fourier based OFDM and found out that the wavelet based
has more advantages than Fourier based OFDM [3]-[6].
In paper [3], descriptions of Fourier based OFDM and
Wavelet based OFDM and differences between them over
the AWGN channel and Rayleigh channel presented. In
paper[7], scatter plot analysis of wavelet OFDM is
presented and symbol reconstruction is shown. In this
paper, wavelet denoising applied on received symbols on
AWGN channel. The transmitted and received symbols
are presented for comparison. The results reveal that The
DWT OFDM has potential to produce hardware
complexity because it doesn’t need a Cyclic Prefix and
proposed system gives nearly perfect reconstruction.
The paper is organized as follows: The system model is
described in Section II. Simulation and results are
discussed in Section III. Concluding remarks appear in
Section
IMULATION & DISCUSSION
Simulation has been carried out to transmit symbols over
wavelet based OFDM system over AWGN channels. The
transmitted symbols for different SNR are examined for
AWGN channel.
For performing the simulations, the flow chain shown in
the Figure 1 is developed under a MATLAB 7.9.0
environment. The blocks are implemented by MATLAB
instructions. The following is an overview of procedures
to simulate the DWT-OFDM scheme shown above.
Transmitter
Input signal: PN Sequence (L=210
– 1 =1023Bits)
Bit pattern Generation: Serial Bits is Converted into 4, 6
or 8 Bit Pattern Suitable for 16, 64 or 256QAM
Generating Look-up table: Look-up table contains
complex values corresponding to a bit pattern.
QAM Encoder (16QAM, 64QAM, 256QAM): QAM
constellation points are obtained by LUT method.
IDWT: Wavelet Decomposition using Haar Wavelets.
Time Domain conversion: Frequency domain signal is
converted to time domain using IDWT and extract Real
time signa