Asymptotic estimation of error fraction corrected by binary LDPC code

Pavel Rybin, Victor Zyablov

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

6 Citations (Scopus)

Abstract

This paper considers new lower bound on fraction of guaranteed corrected errors while decoding the same binary low-density parity-check (LDPC) codes with constituent single parity-check (SPC) and Hamming codes using the same iterative low-complex hard-decision algorithm as in previous works of V. Zyablov and M. Pinsker in 1975 and V. Zyablov, R. Johannesson and M. Loncar in 2009. The number of decoding iterations, required to correct the errors, is a logarithmic function of the code length. The fraction of guaranteed correctable errors computed numerically for various choices of LDPC code parameters with constituent SPC and Hamming codes shows that proposed lower bound gives the better results than previously known best lower bounds obtained by V. Zyablov and M. Pinsker in 1975 for Gallager's LDPC codes and A. Barg and A. Mazumrad for Hamming code-based LDPC (H-LDPC) codes in 2011. Some of obtained numerical results are represented at the end of the paper to demonstrate these improvements.

Original languageEnglish
Title of host publication2011 IEEE International Symposium on Information Theory Proceedings, ISIT 2011
Pages351-355
Number of pages5
DOIs
Publication statusPublished - 2011
Externally publishedYes
Event2011 IEEE International Symposium on Information Theory Proceedings, ISIT 2011 - St. Petersburg, Russian Federation
Duration: 31 Jul 20115 Aug 2011

Publication series

NameIEEE International Symposium on Information Theory - Proceedings
ISSN (Print)2157-8104

Conference

Conference2011 IEEE International Symposium on Information Theory Proceedings, ISIT 2011
Country/TerritoryRussian Federation
CitySt. Petersburg
Period31/07/115/08/11

Fingerprint

Dive into the research topics of 'Asymptotic estimation of error fraction corrected by binary LDPC code'. Together they form a unique fingerprint.

Cite this