Decoding Short LDPC Codes via BP-RNN Diversity and Reliability-Based Post-Processing

TL;DR

This paper introduces a novel decoder diversity architecture using BP-RNN models for short LDPC codes, enhanced with OSD post-processing, significantly improving decoding performance and approaching maximum likelihood decoding.

Contribution

It proposes a new BP-RNN based decoder diversity approach specialized for error classes, combined with OSD post-processing, to enhance decoding of short LDPC codes.

Findings

Specialized BP-RNN decoders outperform standard BP with OSD.

Combining multiple BP-RNN decoders with OSD bridges the gap to maximum likelihood decoding.

The approach effectively leverages error class specialization and post-processing for improved decoding.

Abstract

This paper investigates decoder diversity architectures for short low-density parity-check (LDPC) codes, based on recurrent neural network (RNN) models of the belief-propagation (BP) algorithm. We propose a new approach to achieve decoder diversity in the waterfall region, by specializing BP-RNN decoders to specific classes of errors, with absorbing set support. We further combine our approach with an ordered statistics decoding (OSD) post-processing step, which effectively leverages the bit-error rate optimization deriving from the use of the binary cross-entropy loss function. We show that a single specialized BP-RNN decoder combines better than BP with the OSD post-processing step. Moreover, combining OSD post-processing with the diversity brought by the use of multiple BP-RNN decoders, provides an efficient way to bridge the gap to maximum likelihood decoding.