Research on the complexity of two kinds of Euclidean projection algorithm of ADMM decoding for low-density parity-check codes
王钟斐 1赵悦 1王彪1
扫码查看
点击上方二维码区域，可以放大扫码查看
作者信息
1. 宝鸡文理学院数学与信息科学学院,陕西宝鸡 721013
折叠
摘要
目的 理清低密度校验(Low-Density Parity-Check,LDPC)码的交替方向乘子(Alterna-ting Direction Method of Multipliers,ADMM)译码方法中向量到校验多胞体投影操作的复杂度.方法 比较基于割查找的欧几里德投影算法和快速近似欧几里德投影算法2种典型投影算法的投影结果、算法特点、时间复杂度以及基本操作次数.结果与结论 快速近似欧几里德投影算法各种操作的次数低于基于割查找的欧几里德投影算法,表明前者的时间复杂度低于后者.
Abstract
Purposes—To clarify the complexity of vector to check polytope projection operations in the alternating direction method of multipliers(ADMM)decoding method for low-density parity-check(LDPC)codes.Methods—The projection results,algorithm characteristics,time complexity and basic operation times of the two typical projection algorithms,namely the Euclidean projection al-gorithm based on cut search and the fast approximate Euclidean projection algorithm,are compared.Results and Conclusions—The fast approximation Euclidean projection algorithm requires fewer opera-tions than the Euclidean projection algorithm based on cut search,which indicates that the former has a lower time complexity than the latter.
维普
万方数据