Classification-Based Algorithm for Bit-Flipping Decoding of GLDPC Codes Over AWGN Channels

In this letter, an improved bit-flipping (BF) decoder for generalized LDPC over AWGN channels is proposed. It takes the advantage of the fast BF decoding with only initial help from the channel soft information by introducing a classification step at first. Second, it adds an auxiliary bit in messages between the bipartite graph nodes as a tool for increasing the decision reliability and improving the performance. The reliability is presented as a tool for diminishing the effect of the generated trapping sets. The proposed classification threshold is mathematically derived, and the algorithm is revealed to surpass the various BF algorithms concerning the bit error rate performance.