On the performance and numerical stability of soft-decision Reed-Solomon decoding

In this paper we numerically analyze soft-decision Reed-Solomon decoding based on adaptive Belief Propagation (ABP) algorithm in an additive white Gaussian noise (AWGN) and Rayleigh fading environment. We compare modifications of ABP in terms of frame-error-rate performance and decoding latency, while focusing on efficient implementations in software or hardware. Main outcome of the paper is the analysis of quantization effects when ABP is implemented employing fixed-point values.