Implementation of Soft Computing Techniques for Low Density Parity Check Codes Decoders

Abstract

Ever since Gallager came out with Low Density Parity Check Codes, it is extensively used as correcting codes. At the transmitter’s end, data is encoded using LDPC. When the signal is transmitted in channel, Additive White Gaussian Noise (AWGN), Rayleigh and Rician fading results in signal distortion. At the receiver’s end, signal is decoded to correct the errors. The most commonly used decoding techniques are Sum of Product and Maximum Likelihood algorithms. As both these techniques are dependent on probability of bits, accuracy of decoding is affected. Hence it is necessary to develop efficient encoder/decoder. Instead of developing two different units, one for encoder and the other for decoder, it will be efficient if a simple system is used as both encoder and decoder. Hence, the proposed work aims at developing a LDPC encoder and decoder with soft computing techniques. LDPC being a non-linear code, justifies the use of soft computing technique. Back Propagation Network (BPN) is the most widely used network for non-linear applications. In this work, BPN based decoder is developed