Gaussian Approximation for LDPC Codes under Group Shuffled Belief Propagation Decoding

Group shuffled belief propagation (GSBP) algorithm is a modified algorithm to decode LDPC codes. Compared with belief propagation decoding, this algorithm converges faster. In this paper, we analyze the performance of GSBP decoding by means of Gaussian approximation. Under the assumption of Gaussian density, the updating message means during GSBP decoding are derived. The results show that the proposed algorithm can verify the fast convergence property of GSBP decoding effectively.