Generalized constructions, decoding and implementation of LDPC codes

Date
2008
Authors
Wang, Yige
Contributor
Advisor
Department
Instructor
Depositor
Speaker
Researcher
Consultant
Interviewer
Annotator
Journal Title
Journal ISSN
Volume Title
Publisher
Volume
Number/Issue
Starting Page
Ending Page
Alternative Title
Abstract
Low-density parity-check (LDPC) codes have received significant attention due to their near-Shannon-limit error performance. The belief propagation (BP) algorithm is the best-known decoding algorithm for LDPC codes. However it often needs several tens or hundreds of iterations to converge. Although fast convergence BP algorithms have been proposed, there is neither an easy tool to analyze their performance nor a standard VLSI architecture to implement them. Moreover, there exists a trade-off between waterfall (i.e., threshold) and error floor (i.e., minimum pseudo distance) performances for LDPC codes. For generalized LDPC (GLDPC) codes, better minimum distance properties can be achieved but at the expense of a rate loss. This research investigates performance analysis and hardware implementation of LDPC codes using fast convergence algorithms. Closed-form extrinsic mutual information transfer (EXIT) functions for several fast convergence BP decoding algorithms are proposed. It is shown theoretically that these algorithms converge faster than standard BP, while the corresponding thresholds remain unchanged. Since these algorithms are of serial nature, their parallelization without compromising convergence speed and error performance is also investigated. This result is useful in hardware implementation because in practice partly parallel decoding schemes are often used. This result also sheds some light on how to group nodes in order to preserve performance. The VLSI architecture of a particular decoding algorithm for LDPC codes is proposed. This approach can be adopted in many standards to achieve a large throughput. Doubly GLDPC (DGLDPC) codes are proposed to compensate the rate loss of GLDPC codes. EXIT charts are used to analyze DGLDPC codes and EXIT functions for variable component codes are thoroughly discussed. Based on EXIT charts, differential evolution is used to optimize their threshold. Ensemble weight enumerators are generalized to protograph-based DGLDPC codes. Simulation results show that DGLDPC codes constructed based on these results can improve the performance of their LDPC and GLDPC counterparts in both the waterfall region and the error floor region.
Description
Thesis (Ph.D.)--University of Hawaii at Manoa, 2008.
Doubly GLDPC (DGLDPC) codes are proposed to compensate the rate loss of GLDPC codes. EXIT charts are used to analyze DGLDPC codes and EXIT functions for variable component codes are thoroughly discussed. Based on EXIT charts, differential evolution is used to optimize their threshold. Ensemble weight enumerators are generalized to protograph-based DGLDPC codes. Simulation results show that DGLDPC codes constructed based on these results can improve the performance of their LDPC and GLDPC counterparts in both the waterfall region and the error floor region.
Low-density parity-check (LDPC) codes have received significant attention due to their near-Shannon-limit error performance. The belief propagation (BP) algorithm is the best-known decoding algorithm for LDPC codes. However it often needs several tens or hundreds of iterations to converge. Although fast convergence BP algorithms have been proposed, there is neither an easy tool to analyze their performance nor a standard VLSI architecture to implement them. Moreover, there exists a trade-off between waterfall (i.e., threshold) and error floor (i.e., minimum pseudo distance) performances for LDPC codes. For generalized LDPC (GLDPC) codes, better minimum distance properties can be achieved, but at the expense of a rate loss.
This research investigates performance analysis and hardware implementation of LDPC codes using fast convergence algorithms. Closed-form extrinsic mutual information transfer (EXIT) functions for several fast convergence BP decoding algorithms are proposed. It is shown theoretically that these algorithms converge faster than standard BP, while the corresponding thresholds remain unchanged. Since these algorithms are of serial nature, their parallelization without compromising convergence speed and error performance is also investigated. This result is useful in hardware implementation because in practice partly parallel decoding schemes are often used. This result also sheds some light on how to group nodes in order to preserve performance. The VLSI architecture of a particular decoding algorithm for LDPC codes is proposed. This approach can be adopted in many standards to achieve a large throughput.
Includes bibliographical references (leaves 118-127).
Also available by subscription via World Wide Web
127 leaves, bound 29 cm
Keywords
Citation
Extent
Format
Geographic Location
Time Period
Related To
Theses for the degree of Doctor of Philosophy (University of Hawaii at Manoa). Electrical Engineering; no. 5085
Table of Contents
Rights
All UHM dissertations and theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission from the copyright owner.
Rights Holder
Local Contexts
Email libraryada-l@lists.hawaii.edu if you need this content in ADA-compliant format.