Quasi-Cyclic LDPC环优化构造算法

Quasi-Cyclic LDPC Girth Structure Optimization Algorithm

本文研究了准循环LDPC码的一致校验矩阵与其对应Tanner图环结构之间的关系,将局部平均围长优化算法引入优化构造算法中,提出了一种基于PEG算法的准循环扩展LDPC码构造的改进算法。在改进算法中,首先利用PEG算法产生基矩阵,然后基于平均围长最大准则选择循环移位置换矩阵对基矩阵进行准循环扩展,通过围长分布的优化,从而改善了迭代译码的性能。仿真结果表明,在码长相同、码率和度分布近似情况下,与文献[7]中基于PEG算法的准循环扩展构造的码字比较,改进算法构造的码字能够获得更加优异的性能。

Some important relationships between quasi-cyclic LDPC code and parity check matrix are described. Base on these relationships , the local average girth structure optimization algorithm is taken into structure algorithm. A new algorithm for constructing LDPC codes was proposed, that is an improved algorithm based on quasi-PEG cyclic extension. After a basic matrix was construeted with PEG algorithm, circulant permutation matrices , which are selected on the average girth maximum cyclic shift permutation matrix criteria, would replace the ＂1＂ s in the basic matrix, thus the performance of iterative decoding is improved. The simulation results show that in the same conditions of code length , rate, degree distribution and decoding algorithm, the LDPC codes generated by proposed algorithm outperforms algorithm in [ 7 ].