An Iterative Decoding Method Based on Gaussian Approximation for LDPC Coded Modulations

In this paper, we propose an improved iterative demapping/decoding method for short length low-density paritycheck (LDPC) coded bit-interleaved coded modulations (BICM). For short length LDPC, there exist cycles in the bipartite graph, thereby leading to the high correlation among the extrinsic information of bits out from BP decoder. Based on the Gaussian assumption of the extrinsic log-likelihood ratios (LLRs)delivered by BP decoder, we multiply the extrinsic output of the decoder by a scaling factor and then pass it as a priori information to the demodulator for the next iteration. In this way, we can decrease the correlation of coded bits within the symbol and thus reduce the error propagation. Simulation results show that the proposed method can achieve noticeable performance gains for short length LDPC coded BICM-ID with a minor computation complexity increase.