基于等差数列的LDPC码编码器设计

本文提出一种基于等差数列构造LDPC码的新码类,称为D-LDPC码.文中给出了D-LDPC码的 D 矢量和 D 矩阵的定义,提供一个不含4线循环的确定结构的稀疏奇偶校验矩阵 H 的通用结构,提出一种递归形式的D-LDPC码编码器设计算法.D-LDPC码的编码计算复杂度为O(M),低于卷积码的O(N)复杂度;在中、低码长,任意码率时,性能可与卷积码比美,甚至超越卷积码;编码参数设计灵活,既能与现有标准兼容,又能满足未来发展的需求.

A Designing Method of LDPC Encoder Based on Algorithm Sequence

This paper defines a D-vector and D-matrix,based on which a new ensemble of LDPC code,called as D-LDPC codes,is developed.We discover that the D-matrix has the characteristics of orthogonality and circular-shift,by means of which we construct sparse parity-check matrix based on algorithm sequence.Studying shows that the D-LDPC codes has many advantages,such as the performance of the D-LDPC code is as same as,or in some extent,much better than convolution codes at the condition of middle and low code length and 1/2 code rate;the complexity of the algorithm to encode a D-LDPC code is O(M) which is lower than O(N) of the complexity for the convolution codes;the D-LDPC code has high flexibility concerning design of code length and code rate,which makes it compatible with the existed standard.