Decoding

This chapter provides an overview of efficient soft‐decision decoding techniques. It surveys maximum‐likelihood soft‐decision decoding (ML‐SDD) algorithms, which are relatively simple to apply for short‐length codes. The chapter introduces a novel algorithm based on a softsyndrome decoder, and the interesting topic of SDD algorithms for Reed‐Solomon codes. It focuses on maximum a posteriori probability (MAP) decoding, which underlies the iterative decoding algorithms for random‐like long codes including turbo codes and LDPC codes. Both approaches are considered optimal. However, in MAP decoding the optimality condition minimizes the information bit or symbol error rate, whereas maximum‐likelihood decoding (MLD) minimizes the codeword error rate. The chapter briefly formulates the ML‐SDD of linear codes and reviews several adaptive and fixed reliability‐based MLDand near‐MLD schemes. It introduces some results for rather short packet length and point out open problems for further research.