An improvement of decoding time of ordered statistic decoding for medium length LDPC codes

It is well-known that the Sum-Product Algorithm (SPA) is an efficient algorithm for decoding Low Density Parity Check (LDPC) codes. In spite of this, for LDPC codes with medium length which are now attracting interests for the application to fifth generation systems, it is known that decoding error rate is much inferior to that of Maximum Likelihood Decoding (MLD). In this paper, we improve bit error rate of SPA with so called Ordered Statistic Decoding (OSD) method by Fossorier [1]D Especially, we improve stopping condition of combinatorial search in OSD method and demonstrate by simulation that the proposal stopping condition is more efficient compared to Fossorier (IEEE J. Sel. Areas Commun., 19, 2001). For that purpose, we introduce a conception “coincidence rate” which plays an important role.