Development of Fuzzy Receiver for GSM Application

Abstract

A channel equalizer is one of the most important subsystems in any cellular
communication receiver. It is also the subsystem that consumes maximum computation
time in the receiver. Traditionally maximum-likelihood sequence estimation (MLSE) was
the most popular form of equalizer. Owing to non-stationary characteristics of the
communication channel MLSE receivers perform poorly. Under these circumstances
maximum-aposteriori probability (MAP) receivers also called Bayesian receivers perform
better. This thesis proposes a fuzzy receiver that implements MAP equalizer and provides
a performance close to the optimal Bayesian receiver. Here Bit Error Rate (BER) has
been used as the performance index.
This thesis presents an investigation on design of fuzzy based receivers for GSM
application. Extensive simulation studies which shows that the performance of the
proposed receiver is close to optimal receiver for variety of channel conditions in
different receiver speeds where channel suffers from Rayleigh fading. The proposed
receiver also provides near optimal performance when channel suffers from
nonlinearities.