一种提高LDPC译码器吞吐率的译码算法  

A Higher Throughput LDPC Decoding Algorithm

作　　者：张金贵[1] 斐文端[1] 许星辰[1] 姜文哲 

机构地区：[1]中国电子科技集团公司第五十四研究所,河北石家庄050081 [2]济南高新区管委会,山东济南250101

出　　处：《无线电工程》2008年第6期49-52,共4页Radio Engineering

摘　　要：为了设计高效的LDPC译码器,结合准循环结构LDPC的校验矩阵H的规律性、乘性修正最小和译码算法不需要估计信道质量的特点和部分并行译码实现复杂度低的特点,介绍了一种新的译码算法——交迭的部分并行译码算法,这种译码算法相对于采用部分并行结构的BP译码算法,不但降低了硬件实现的复杂度,减少了存储资源的开销,而且提高了译码器的吞吐率。In order to design an efficient decoder of LDPC, combining the regularity of check matrix H of Quasi-cyclic LDPC, the characteristic of NMSA （normal min-sum algorithm） that don＇t have to evaluate the quality of the channel and the characteristic of partially parallel decoding that the implementing complexity is low, we described a new decoding algorithm, overlapped partially parallel decoding algorithm, which can not only lower the complexity of implementation and the overhead of memory but also improve the throughput of decoder in comnarison with decoding algorithm of nartiallv narallel.

关 键 词：乘性修正最小和译码算法 部分并行译码 交迭的部分并行译码 吞吐率 

分 类 号：TN911[电子电信—通信与信息系统]