Performance of product codes on channels with memory

Abstract

In this paper, we present an analytical method to compute the performance of product codes on channels with memory. A channel with memory introduces correlated error bursts, and hence, it is described as hidden Markov models. A hard-decision distance-bounded decoder for product code is assumed. Such an analysis is relevant for evaluating the performance of product-code-based error-correcting scheme employed in conventional tape drive systems. Using this analytical method, we present codeword-error rate plots for various Reed-Solomon codes based product codes.