Research On Construction Algorithm Of Bipartite Graph For LDPC Codes

Low-density parity-check codes (LDPC) in graphical models come to be one of the best coding technologies because of their low-complicity iterative decoding algorithm and capacity approaching performance, and it is how to construct LDPC code with good performance that attract more and more researchers'eyes in recent years. The construction of bipartite graph plays an important role in the LDPC code-construction which is based on graphical models. The performance of low-density parity-check codes can be improved by constructing a good Tanner graph.In this dissertation, the basic principles of the coding and decoding of LDPC codes are studied systematically. A new algorithm of bipartite-graph-expanding is proposed for construction of Tanner graph. The main results and contents are summarized as follows.1. The principles of coding and decoding for low-density parity-check codes are briefly summarized, the main factors which can impact the performance of LDPC codes such as degree sequence, cycles and connectivity of Tanner graph are systematically analyzed, emphasis on the graph connectivity.2. Several algorithms for constructing Tanner graph of LDPC codes such as cycle-eliminating algorithms and connectivity-improving algorithms are analyzed in detail. The simulation results show that the codes constructed with these algorithms has lower bit error rate and error floor. The graph of simulation performance is presented. Besides, the performance of the codes constructed with these algorithms is analyzed.3. Based on the study of construction algorithm mentioned before, a mothed of computing ACE value is given, a new bipartite-graph-expanding algorithm based on EMD algorithm is proposed, and the new algorithm is fit for EMD algorithm. Moreover, the software simulation was given, and the performance of EMD algorithm was analyzed.