Study Of Ldpc Codes, Encoding And Decoding Algorithms

Low-density parity-check (LDPC) codes are one kind of linear block codes approaching Shannon limit based on sparse parity-check matrices. They are with excellent error correcting ability, flexibility, and low decoding complexity. So they are becoming a power competitor in the coding research field, and have been successfully applied in various fields of communications recently.In this paper, the encoding and decoding algorithms for LDPC codes are mainly studied. Firstly, the communications and channel coding theory are introduced; secondly, the properties, developments and applications of LDPC codes are presented; thirdly, the random constructions and structured constructions for parity-check matrices of LDPC codes are analyzed. Lastly, the encoding and decoding issues are studied. For the BP decoding algorithms, several main factors affecting the decoding performance are analyzed. And then, four simplified and one improved BP decoding algorithms for LDPC codes are proposed, which are analyzed and compared in detail with MATLAB simulation. Compared with simplified algorithms, the improved BP decoding algorithm has better performance. For the encoding issue, three encoding algorithms for LDPC codes and their complexities are studied. Especially, a linear complexity encoding method is investigated based on Tanner graphs for arbitrary LDPC codes and its performance is analyzed.