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, 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. We show that the OSD post-processing step effectively takes advantage of the bit-error rate optimization, deriving from the use of binary cross-entropy loss function, and the diversity brought by the use of multiple BP-RNN decoders, thus providing an efficient way to bridge the gap to maximum likelihood decoding.
翻译:本文根据信仰-传播算法的经常性神经网络模型,对低密度对等检查(LDPC)短时间代码的解码多样性结构进行了调查。我们提出了实现解码多样性的新办法,将BP-RNN解码器专门用于具体的错误类别,并为其提供吸收支持。我们进一步将我们的方法与订购的统计解码(OSD)后处理步骤结合起来。我们表明,OSD后处理步骤有效地利用了因使用二元交叉热带损失功能而产生的位拉速优化,以及多种BP-RNND解码器带来的多样性,从而为弥合差距以最大限度地解码提供了有效的途径。