Chinese Journal of Computers   Full Text
Title	Performance Impact of Deferent Permutation Matrix on LDPC Codes Based on Partitioned H-Matrix
Authors	PENG Li ZHU Guang-Xi
Address	(Department of Electronics and Information Engineering， Huazhong University of Science and Technology, Wuhan 430074)
Year	2008
Issue	No.5(783—792)
Abstract &   Background	Abstract The authors investigate three kinds of circulant-shift permutation matrices, which are called I (identity matrix), D and Q matrices, and discuss their structural characteristics and several properties in this paper. Computer simulating tests show that D and Q matrices can improve performance of the regular QC-LDPC codes based on identity matrix, including irregular QC-LDPC codes in IEEE 802.16e standard draft. The other contribution in the paper is designing the new family of S-LDPC codes by using Q matrix as partitioned matrices of H-matrix, which have the advantage of irregular QC-LDPC codes presented in IEEEE 802.16e standard draft in performance and calculating complexity. Keywords low-density parity-check codes; encoder; parity-check matrix; circulant-shift permutation matrix; recursive encoder Background Low-density parity-check (LDPC) codes belong to a class of linear block code in error-correcting code fields. The LDPC codes, discovered by Gallager in early 1960’s, were rediscovered in the late 1990s, and have lately attracted a great deal of attention because of showing Shannon-limit-approaching performance and low-complexity parallel decoding architecture. Since their rediscovery, design, construction, theoretical analyses and applications of these codes have become focal points of research. Especially, code constructions and linear-complexity encoding algorithms were paid attention in error-correcting code fields. Because of breakthrough development of linear encoding algorithm, the encoding and decoding systems of several classes of LDPC codes are introduced in industry standard.