LDPC码数据分配通用模块设计方案  

Method for general-purpose module of data distribution of LDPC code

作　　者：王健[1] 李集林[1] 战勇杰[1] 

机构地区：[1]北京卫星信息工程研究所,北京100083

出　　处：《电子设计工程》2012年第20期1-4,共4页Electronic Design Engineering

基　　金：863计划(国家高技术研究发展计划)课题(2009AA12Z149)

摘　　要：上世纪60年代初,香农的学生Gallager在他的博士毕业论文中首次提出了LDPC码的概念和完整的译码方法,但是直到上世纪末期,随着LDPC码译码理论的进步和计算机技术的发展,LDPC码才以其优良的误码性能和良好的可实现性成为人们研究的焦点。针对QC类LDPC码进行研究的时候,注意到很多码的循环子矩阵中不只有一组1,这就产生了水平运算后判断运算结果属于哪个存储模块的问题;另外,由于校验矩阵中每个循环子矩阵的列初始位置都是不同的,而且通常LDPC码的校验矩阵的循环子矩阵的数目都是非常庞大的,因此如果通过程序固化的方法,不但容易出现不易排查的错误,而且开发效率会大大下降。为解决该问题,并将解决方案通用化,文中以校验矩阵中循环子矩阵中1的排列特点为研究对象,找到引起变化的量之间的共性特征,从而实现通用化模块的设计。At the beginning of 60 s of last century,Gallager,the student of Shannon issued the concept and a whole set of method for decoding of LDPC code at the first time in the world. But the LDPC code had not been on the focus of our research with its super performance until the end of the last century, for the improvement of the LDPC decoding theory and the development of the computer technology. When we research the QC-LDPC code, we realize that many kinds of LDPC codes do not only have one set of 1 ,and this leads the question that the result of CNU belongs to which memory module. Meanwhile, because of the difference of all column initial position of every submatrix and the large number of the submatrix of the check matrix, we can not easily solidify the location of each 1 and this will lead errors which can not be easily found and the low efficiency of exploitation. To solve this question and generalize the module, this thesis researches the character of the arrangement rules of lin submatrix, finds the commonness of variable, and realizes the design of the general-purpose module.

关 键 词：LDPC 通用 准循环 数据分配 

分 类 号：TN919[电子电信—通信与信息系统]