Research On Decoding Algorithm And Combating Burst Error Performance Of LDPC Codes

Both decoding algorithm optimization and combating burst error scheme of LDPC codes are investigated in this dissertation.In order to tradeoff the decoding performance and complexity of low-density parity-check (LDPC) codes, joint probability of weighted bit-flipping decoding algorithm is proposed. Firstly, utilize the initial probability value of channel to get hard decision and then calculate the syndrome. Secondly, find out the nodes that don’t satisfy the check equation and according to node messages processing approach of belief propagation (BP) algorithm to update local node messages. Finally, the variable nodes collect all related messages and take these as soft information to calculate weighted error term, which makes the location of the flipped bit more reliable. Simulation results show that compared with typical algorithms, the new algorithm can improve more than1.0dB on decoding performance while maintaining a relative low decoding complexity, so it is a more practical decoding algorithm for LDPC codes.For the purpose of degradating the severe influence on decoding performance by the existence of girth especially small girth of given check matrix, taking the distance property into account, an simple but practical burst error detector is designed. In the first, calculate the posterior probability value and make hard decision. And then, generate parity flag based on syndromes got from check matric. At the end, obtain burst error flag accounting to designed standard and accurately locate the length and position of burst errors. Simulation results indicate that the burst error detector can effectively improve the capability to correct block...