Iterative decoding of cyclic codes based on FFT sliding correlator

Abstract

In this dissertation, our objective is to propose an alternative way of implementing MLD for cyclic block code which might offer better decoding efficiency with relatively fast processing. We successfully designed the soft-in-hard-out ML-decoding of cyclic code based on sliding correlator. BER performance results indicate that our decoder performs well as it meets the theoretical expectation by having performance curves which are slightly better than the theoretical upper bound by 0.3-0.4dB for various BCH codes. Furthermore, FFT implementation for the sliding correlator contributes to the fast and efficient processing of our decoder.