Linear Complexity Lossy Compressor for Binary Redundant Memoryless Sources

TL;DR

This paper introduces a linear complexity lossy compression algorithm for binary redundant memoryless sources using sparse graph codes and extended belief propagation, achieving near-optimal performance.

Contribution

It presents a novel linear complexity compressor based on extended belief propagation with an inertia term for binary redundant memoryless sources.

Findings

Achieves near-optimal compression performance for moderate block lengths.

Uses a nonlinear function within sparse graph codes for redundancy exploitation.

Demonstrates the effectiveness of the proposed method through simulations.

Abstract

A lossy compression algorithm for binary redundant memoryless sources is presented. The proposed scheme is based on sparse graph codes. By introducing a nonlinear function, redundant memoryless sequences can be compressed. We propose a linear complexity compressor based on the extended belief propagation, into which an inertia term is heuristically introduced, and show that it has near-optimal performance for moderate block lengths.