Iterative Decoding and Estimation of Two-Dimensional Channels with Memory

In this paper we consider two-dimensional channels with memory and with error patterns modeled as a Markov random field. We develop an joint iterative estimation and decoding algorithm by using output from an LDPC decoder in combination with an error estimation algorithm based on techniques from statistical physics and digital image restoration to find an estimate of the error pattern and use this estimate then in turn to provide emph{a priori} information for use in the decoder. Simulations show that this approach can significantly reduce the occurrence of bit errors compared with standard iterative decoding techniques.