Rereach On BICM-ID Systerms

Joint Low-Density Parity-Check (LDPC) coded Bit Interleaved Coded Modulation with Iterative Decoding (BICM-ID) technology, due to the high spectrum efficiency, structure flexibility, low implementation complexity and decoding performance advantages, has been widely concerned. In a LDPC coded BICM-ID system, LDPC codes with different distributions is of the large difference in performance, by the best of the different modulation methods and corresponding distribution of LDPC codes can greatly improve system performance.Irregular LDPC code has better error floor falls domain than regular LDPC code, and for a particular modulation scheme, there is a certain degree distribution of an LDPC code of the lowest bit error rate of the communication system. Extrinsic information transfer chart (EXIT) is an effective method for the analysis of the convergence of iterative decoding, take advantage of EXIT chart to predict convergent and critical SNR of different degree distribution of LDPC codes under different modulation schemes. In this paper, EXIT chart analyze is used, the combination of coding and modulation in the BICM-ID system to optimize the degree distribution of the LDPC code, and based on PEG algorithm and ACE algorithm, The PEG-ACE algorithm is proposed to construct LDPC code, simulation shows that a slight loss of performance has great improved the error floorIn BICM systems, Gray mapping is the optimal mapping, but in BICM-ID systems, Gray mapping isn’t the optimal mapping since it has no iteration gain, we propose a16QAM constellation which matches BICM-ID system with the LDPC encoding. Simulation results show that this mapping has0.1dB gain over Gray mapping, and EXIT chart curve by the verified constellation mapping thesis correctness of our design.Based on the Sum-Product and Min-Sum decoding algorithm, we propose an improved min-Sum decoding algorithm, the improved algorithm not only can reduce the amount of computation and storage overhead, but also can improve the LDPC decoding speed. But when the coded length is large and degree distribution is complexity. Whether traditional or improved decoding algorithm, the decoding complexity and the storage space is still large, hardware implementation is difficult. In this paper, we propose the layered decoding algorithm For dual-diagonal structure of quasi-cyclic LDPC codes, this decoding algorithm has a slight performance loss compared to the several decoding algorithm, but it can greatly reduce the complexity of hardware implementation, has good practicability.