Memory-Assisted Quantized LDPC Decoding

TL;DR

This paper proposes a memory-assisted approach for quantized LDPC decoding that reuses check node messages across iterations, significantly improving decoding performance and efficiency under coarse quantization.

Contribution

It introduces a memory reuse technique for LDPC decoding that mitigates information loss from coarse quantization, along with an optimized quantization method considering message statistics.

Findings

Performance improved by up to 0.23 dB with memory assistance.

Memory reuse enables up to 32% better area efficiency for 2-bit decoding.

The approach reduces memory requirements through a simple merge operation.

Abstract

We enhance coarsely quantized LDPC decoding by reusing computed check node messages from previous iterations. Typically, variable and check nodes update and replace old messages every iteration. We show that, under coarse quantization, discarding old messages entails a significant loss of mutual information. The loss is avoided with additional memory, improving performance by up to 0.23 dB. We optimize quantization with a modified information bottleneck algorithm that considers the statistics of old messages. A simple merge operation reduces memory requirements. Depending on channel conditions and code rate, memory assistance enables up to 32 % better area efficiency for 2-bit decoding.