Joint Optimization of Protograph LDPC Code Pair for Joint Source and Channel Coding

Chen Chen, Lin Wang, Chung Ming Lau

Research output: Journal article publicationJournal articleAcademic researchpeer-review

20 Citations (Scopus)

Abstract

Double protograph low-density parity-check code has been shown to be a good solution for joint source-channel coding. Moreover, the source code and channel code (as one code pair) can be jointly optimized to achieve a good bit error rate (BER) performance. In this paper, first, an optimal design criterion of this code pair based on an extended curve-fitting algorithm is proposed. Second, a joint optimization procedure which searches for both the optimal decoding threshold and the optimal code pair is presented. In addition, for short-to-moderate-length codewords, the finite-length effect cannot be ignored and the procedure has to be modified with a new notion named length-N decoding threshold. The protograph extrinsic information transfer analysis and the BER simulation have shown that our optimized code pairs acquire lower decoding thresholds than conventional code pairs. Moreover, the optimized code pairs with short-to-moderate-length codewords can mitigate the finite-length effect and obtain low length-N decoding thresholds.

Original languageEnglish
Article number8314100
Pages (from-to)3255-3267
Number of pages13
JournalIEEE Transactions on Communications
Volume66
Issue number8
DOIs
Publication statusPublished - 1 Aug 2018

Keywords

  • and protograph LDPC
  • Joint optimization
  • length-N decoding threshold
  • optimal design criterion

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this