Research On Construction Of LDPC Codes And Encoding And Decoding Algorithm In CMMB

Low-Density Parity-Check Code (LDPC) is a kind of rapidly developed linear blockcodes in recent years. It has the performance approaching to Shannon limit in use ofiterative decoding algorithms as Believe Propagation (BP) Algorithm. The LDPC code hasmany advantages just like low complexity, high encoding gain, and flexible bit-ratestructure and so on. At the same time, it can be achieved by parallel decoding structure.Therefore, LDPC code is one of the most talked about topics in the field of channel coding.In this paper, the highly structured LDPC code which is defined in China mobilemultimedia broadcasting system (CMMB) standard is introduced in detail. The features ofthe check-up matrix of LDPC code are similar to the structure of the quasi-cyclic matrix.For this reason, an indirect way to construct the check-up matrix of the LDPC codes in theCMMB standard is proposed using the structural characteristics and the encoding method ofthe quasi-cyclic LDPC (QC-LDPC) code. Firstly, a standard matrix with regularquasi-circular structure is constructed and encoded based on the cycle characteristics. Then,the obtained code word is converted into the output form which the system needs accordingto the transformation rules. It can not only reduce the complexity of the encoding algorithmbut also decrease the amount of the storage space in the process.In addition, the decoding algorithms of LDPC are introduced in details in this paper.Especially the basic model and various kinds of the improved methods of BP algorithm areintroduced, such as Min Sum-Product Algorithm (MSA) and Layered BP (LBP). Animproved plan of the existing LBP algorithm is given according to the structurecharacteristics of LDPC code in the system of CMMB. The check-up matrix ofCMMB-LDPC is processed in block to complete decoding by the serial mode combinedwith the parallel mode so as to improve the computational efficiency. At last, in theconditions of Additive White Gaussian Noise (AWGN) channel, BPSK modulation, the same signal-to-noise ratio (SNR) and different iteration times, the computer simulation andthe analysis of bit error performance of the various decoding algorithms are provided.