摘要
针对一类规则QC-LDPC码,提出一种高效的Log-BP译码方法,通过矩阵分裂,将原监督矩阵分裂成多个小矩阵,将原本的校验节点更新运算拆分成多次处理,以降低log-BP迭代运算的复杂度,给出该方法的迭代运算顺序。与现有的log-BP译码方法相比,该方法在相同的码速率下,校验节点运算单元与变量节点运算单元总规模减小了1/3;在相同的硬件资源下,译码速率提高了1/3,且校验节点运算单元与变量节点运算单元结构趋于对称,有利于设置更少的流水线级数、获得更好的时钟性能。
This paper presents a decoding method for regular QC-LDPC codes. In this method, the original check matrix is split into several smaller ones, thus CNU is decomposed into multiple parallel units, which results in great hardware resource reduction. A novel iterative sequence is presented. Compared with the traditional log-BP decode method, the method can reduce the whole logic core size of CNU and VNU by 1/3 under the same bit-rate, and it can improve bit-rate by 1/3 under the same logic core size in the condition of the original check matrix split into two parts.
出处
《计算机工程》
CAS
CSCD
北大核心
2008年第14期9-11,共3页
Computer Engineering
