Uniform vs. Non-Uniform Coarse Quantization in Mutual Information Maximizing LDPC Decoding

TL;DR

This paper demonstrates that uniform quantization in LDPC decoders, despite being simpler, causes minimal mutual information loss and only slight performance degradation, offering a more efficient decoding approach.

Contribution

It introduces a uniform quantization method for LDPC decoders that reduces complexity and delay with negligible performance loss, compared to traditional non-uniform quantization.

Findings

Uniform quantization causes less than 0.01 dB performance loss.

Decoders with 3-bit uniform quantization outperform traditional 4-bit decoders in complexity.

Mutual information loss from uniform quantization is minimal.

Abstract

Recently, low-resolution LDPC decoders have been introduced that perform mutual information maximizing signal processing. However, the optimal quantization in variable and check nodes requires expensive non-uniform operations. Instead, we propose to use uniform quantization with a simple hardware structure, which reduces the complexity of individual node operations approximately by half and shortens the decoding delay significantly. Our analysis shows that the loss of preserved mutual information resulting from restriction to uniform quantization is very small. Furthermore, the error rate simulations with regular LDPC codes confirm that the uniformly quantized decoders cause only minor performance degradation within 0.01 dB compared to the non-uniform alternatives. Due to the complexity reduction, especially the proposed 3-bit decoder is a promising candidate to replace 4-bit conventional decoders.