LDPC码的一种低复杂度译码算法及关键电路设计  

Low complexity decoding algorithm for LDPC codes and design of key circuits

作　　者：马克祥 刘毅[1] 胡建华[1] 孙吉成[1] 张海林[1] 

机构地区：[1]西安电子科技大学综合业务网理论及关键技术国家重点实验室,陕西西安710071

出　　处：《西安电子科技大学学报》2013年第6期6-12,共7页Journal of Xidian University

基　　金：国家自然科学基金资助项目(61072069);教育部科学技术研究重点(重大)资助项目(2010ZX03002-005);高等学校学科创新引智计划资助项目(B08038);西安电子科技大学自主科研资助项目(72001859)

摘　　要：并行加权比特翻转算法(PWBF)可以获得好的译码性能,但其比特选择机制计算较为复杂,不利于硬件实现.通过对PWBF算法比特选择机制的改进,提出一种低复杂度的低密度奇偶校验码(LDPC)译码算法.具体来讲,每次迭代过程中,当完成所有比特的品质因素更新后,挑选品质因素最大的若干比特进行翻转译码.另外,笔者对算法关键模块的硬件实现进行分析,分别给出了计算优化的电路结构设计.与PWBF算法相比,笔者提出的算法和优化技术大大降低了LDPC译码器的复杂度.Parallel Weighted Bit Flipping（PWBF） can achieve a good decoding performance. However, it is hard for the hardware design and implementation because of the high complexity of its bit-chosen mechanism. By improving the bit-chosen mechanism in PWBF, a low-complexity decoding algorithm is proposed in this paper. Especially, in each iteration step of decoding, after the metric value of every bit is updated, several bits with the largest metric values are flipped. Furthermore, the optimized circuits with low complexity are provided for the critical modules of the proposed algorithm. Compared with the PWBF algorithm, the complexity of LDPC decoders is greatly decreased by use of the proposed algorithm and the optimized circuits.

关 键 词：低密度奇偶校验码 加权比特翻转算法 并行二叉树 选择网络 

分 类 号：TN911.22[电子电信—通信与信息系统]