Iterative row-column soft-decision feedback algorithm using joint extrinsic information for two-dimensional intersymbol interference

In this paper, we redesign the previous iterative row-column soft decision feedback algorithm (Cheng et al., 2007) for a two-dimensional intersymbol interference channel with additive white Gaussian noise; a 2 × 2 averaging mask and two 3 × 3 masks are considered. In particular, we consider the joint statistics for a block of pixels involved in the exchange of extrinsic information between the detectors; previously, these pixels were considered to be statistically independent. The new algorithm, referred to as the block (BLK) algorithm, provides more than 1 dB SNR gain; the resulting performance is within 0.3 dB from the maximum likelihood bound for the masks considered. To address the increased computational and storage complexity introduced by the joint statistics, we have developed a simplified version of the block (SBLK) algorithm. Experimental results demonstrate that the SBLK algorithm performs almost as well as the BLK algorithm.