A Parallel Layered Decoding Algorithm for LDPC Codes in WiMax System

WiMax presents a quasi-cyclic LDPC codes with excellent performance in decoding, which is suitable to parallel decoding for its structure. And the flexible matrix structure can support more implementation of code with different code-length. To decrease the iterative times of the decoding algorithm and save the hardware resource of the decoder for LDPC codes, the paper presents a parallel layered decoding algorithm based on OMS, which doesn´t reduce the decoding performance. This algorithm proposes a novel scheduling to improve the updating of message. Simulation results show that, for a 2304 bits length of code, the parallel layered decoding algorithm exhibits a better convergent performance than a traditional algorithm. The SNR of parallel layered OMS decreases about 0.05 dB than that of traditional OMS, when the BER is 10^-3.