Search for Good Irregular Low-Density Parity-Check Codes Via Graph Spectrum

Research on the expander code shows that for a regular low-density parity-check (LDPC) code, the Tanner graph’s spectrum determines its properties, such as the minimum distance and the size of stopping sets. In this study, we demonstrate theoretically and experimentally that the performance of irregular LDPC codes is related to the graph spectrum. Our observations may provide an efficient metric to search for good irregular LDPC codes.