基于迭代译码的LDPC码稀疏校验矩阵重建

为了改善高误码率情况下低密度奇偶校验（LDPC）码稀疏校验矩阵重建算法的性能，基于迭代译码的思想提出了一种稀疏校验矩阵的重建算法。首先，利用对偶空间算法获取到部分非稀疏校验向量，并对其进行稀疏化处理。其次，利用稀疏化后的校验向量对LDPC码进行软判决迭代译码，从而对码字中错误比特进行纠正，以改善码字质量。然后，对纠错后码字再次进行校验向量获取，不断重复迭代。最后，实现LDPC码稀疏校验矩阵的重建。实验结果表明:在误码率为10-3量级下，针对IEEE802.16e、IEEE802.11n等协议下的LDPC码，所提算法均能有效完成重建，同时新算法的稀疏矩阵重建率要明显好于传统方法。 

Reconstruction of sparse parity-check matrix of LDPC codes based on iterative decoding

In order to improve the performance of Low Density Parity Check (LDPC) code sparse parity-check matrix reconstruction algorithm under high error rate, this paper proposes a sparse parity-check matrix reconstruction algorithm based on the idea of iterative decoding. First, the dual space algorithm is used to obtain part of non-sparse check vector and make it sparse. Second, the sparse check vector is used to perform soft-decision iterative decoding of the LDPC code, thereby correcting the error bits in the codeword to improve the quality of the codeword. Then, the check vector is obtained again for the codeword after error correction. Finally, with repeated iteration, the reconstruction of the sparse parity-check matrix of the LDPC code is realized. The experimental results show that the algorithm in this paper can effectively complete the reconstruction of LDPC codes under IEEE802.16e, IEEE802.11n and other protocols at a bit error rate of 10-3, and the sparse parity-check matrix reconstruction rate of the proposed algorithm is significantly better than that of traditional methods.