Modified LDPC BF decoding algorithm based on reliability ratio

Compared with belief-propagation decoding algorithm which has complexity class of o(n2), bit-flipping (BF) decoding algorithm for low density parity check (LDPC) codes has only complexity class of o(n). However, the performance of BF algorithm is degraded. This paper presents a modified BF decoding algorithm, which uses reliability ratio as a weight of information node in unsatisfied parity checks so that the soft information is introduced in decoding, and the decoding performance is improved greatly with a little more computation. Analyses show the modified algorithm still has a complexity class of o(n). Simulation results indicate that the error rate with 7 dB SNR is improved from 10-3 of the weighted bit-flipping decoding algorithm to 10-4 of the modified algorithm.