大围长结构化LDPC码构造方法

在LDPC码的构造中,校验矩阵拥有大围长对于改善码的性能有着重要的意义.结构化是提高码实用性的关键.提出一种低复杂度的基于列差搜索法(Column-Difference Search AIgorithm)的准循环LDPC码构造方法,用以设计大围长和任意码率的规则QC-LDPC码(以下称为CDS-LDPC码).该方法可线性编码,易于IC实现.仿真结果表明,该方法构造的各种码率CDS-LDPC码在BER性能上均优于对应的随机码,与同属QC-LDPC码的Tanner码和Array码相比明显提高了0.79～3.28dB,并且在码长、码率等参数设计上更为灵活.

Design of Structured LDPC Codes with Large Girth

A parity-check matrix H with large girth has important significance to improve the performance of LDPC codes. And the key to the encoder implementation is the algebraic code structure. This paper proposed a novel code construction algorithm with low complexity based on the Column-Difference Scarch(CDS) Algorithm, which can design regular Quasi Cyclic LDPC codes with large girth and arbitrary code rate. It has linear encoding complexity and is friendly to hardware implementation. The experimental results show that CDS-LDPC codes with different code rates perform better than Tanner codes and Array codes,which increase 0. 79--3. 28dB than another two classical QC-LDPC codes,and also outperform the counterparts of random codes. In addition,CDS-LDPC codes have more flexibility on the design of code length and rate.