Multiple Hypotheses Iterative Decoding of LDPC in the Presence of Strong Phase Noise

TL;DR

This paper introduces a multiple hypotheses iterative decoding algorithm for LDPC codes that effectively handles strong phase noise in satellite communication systems, improving detection accuracy and reducing pilot symbol requirements.

Contribution

It presents a novel multiple hypotheses decoding approach combined with a low complexity mixture reduction for joint phase noise estimation and data detection.

Findings

Enhanced decoding performance in high phase noise conditions

Reduced pilot symbol requirements compared to traditional methods

Maintains high accuracy with low complexity mixture reduction

Abstract

Many satellite communication systems operating today employ low cost upconverters or downconverters which create phase noise. This noise can severely limit the information rate of the system and pose a serious challenge for the detection systems. Moreover, simple solutions for phase noise tracking such as PLL either require low phase noise or otherwise require many pilot symbols which reduce the effective data rate. In the last decade we have witnessed a significant amount of research done on joint estimation and decoding of phase noise and coded information. These algorithms are based on the factor graph representation of the joint posterior distribution. The framework proposed in [5], allows the design of efficient message passing algorithms which incorporate both the code graph and the channel graph. The use of LDPC or Turbo decoders, as part of iterative message passing schemes, allows the receiver to operate in low SNR regions while requiring less pilot symbols. In this paper we propose a multiple hypotheses algorithm for joint detection and estimation of coded information in a strong phase noise channel. We also present a low complexity mixture reduction procedure which maintains very good accuracy for the belief propagation messages.