论文部分内容阅读
不同于Turbo码,LDPC是数学结构非常好的一类高速码。对其结构进行系统研究,不但对实际编码方案很有帮助,而且也有很高的理论价值。二进制低密度码(MN码)编码的基础为Gaussian消元法(消去法)。文中推导了MN码的一致校验矩阵的基于初等变换的Gaussian消元法。研究表明,基于初等变换的Gaussian消元法结合“minprod”算法在进行LU分解时,可以有效地保持MN码一致校验矩阵的稀疏性。
Unlike Turbo codes, LDPC is a very good class of high-speed codes in mathematical structures. Systematic research on its structure not only helps the actual coding scheme, but also has high theoretical value. Binary low-density code (MN code) encoding is based on the Gaussian elimination method (elimination method). In this paper, the Gaussian elimination method based on the elementary transformation of the uniform parity check matrix of MN codes is derived. The results show that Gaussian elimination based on elementary transformation combined with “minprod” algorithm can effectively preserve the sparseness of MN parity check matrix in LU decomposition.