LDPC codes: comparing cluster graphs to factor graphs

TL;DR

This paper compares cluster graph and factor graph representations of LDPC codes, demonstrating that cluster graphs offer advantages in computational efficiency, convergence speed, and accuracy over factor graphs.

Contribution

It provides a novel comparison showing that cluster graph representations outperform factor graphs for LDPC code inference tasks.

Findings

Cluster graphs improve inference accuracy for LDPC codes.

Cluster graphs reduce computational cost compared to factor graphs.

Cluster graphs enhance convergence speed in decoding LDPC codes.

Abstract

We present a comparison study between a cluster and factor graph representation of LDPC codes. In probabilistic graphical models, cluster graphs retain useful dependence between random variables during inference, which are advantageous in terms of computational cost, convergence speed, and accuracy of marginal probabilities. This study investigates these benefits in the context of LDPC codes and shows that a cluster graph representation outperforms the traditional factor graph representation.