具有低编码复杂度准循环扩展LDPC码的构造方法

摘要

PEG(Progressive-Edge-Growth)算法是迄今为止构造性能优异的LDPC中短码的一种有效构造方法,然而直接采用该算法构造的LDPC码的编码复杂度正比于码长的平方,这是其实用化过程中的一个瓶颈.针对这一问题,提出一种具有低编码复杂度和低错误平层的准循环扩展LDPC码的构造方法.该算法在PEG算法基础上,先构造出近似下三角结构的半随机基矩阵,然后再对基矩阵进行扩展,该方法可以在不改变基矩阵的度分布比例情况下,有效消除短环.仿真结果表明,所提出的方法构造的LDPC码比原始的PEG算法构造的随机LDPC码具有更低的错误平层,而且编码复杂度更低,更易于硬件实现.%Progressive-Edge-Growth (PEG) algorithm is an efficient method for constructing LDPC codes with short and intermediate block lengths. However, the complexity of codes produced directly by this algorithm is proportional to quadratic length of the codes, which restricts the implementation of the LDPC codes. To solve the problem, a method was proposed for constructing quasi-cycle extension LDPC codes with low encoding complexity and good error performance. Based on PEG algorithm, a base matrix with an approximate lower triangular was constructed, then all its elements were replaced by circulation permutations. The new method could eliminate the short cycles effectively without changing the degree distribution fraction of the basic matrix. The simulation results show that; compared with LDPC codes generated by PEG algorithm, the LDPC codes presented here have lower encoding complexity and better error performance, which is easy for implementation with hardware.