基于能量的稀疏自适应匹配追踪算法  

ENERGY-BASED SPARSITY ADAPTIVE MATCHING PURSUIT ALGORITHM

作　　者：李保珠[1] 邵建华[1] 聂梦雅[1] 刘刚[1] 

机构地区：[1]南京师范大学,江苏南京210023

出　　处：《计算机应用与软件》2015年第11期121-125,共5页Computer Applications and Software

基　　金：教育部博士点基金课题新教师类项目(2013102SBJ0265)

摘　　要：信号的重建算法在整个压缩感知领域中居于重要的地位。针对稀疏度未知的情况下的信号重建,在经典的稀疏自适应匹配追踪(SAMP)算法的基础上,提出一种基于能量的稀疏自适应匹配追踪(ESAMP)算法。根据测量向量与重建信号能量的比值自适应调整步长,确定步长的合理初始值,对二进制信号的重建算法进行进一步修正,提高了二进制信号的重建精度并实现了二进制信号的完整重建。仿真结果表明,在相同条件下该算法能够在提高重建速度的同时保证较高的重建精度,以更优越的综合性能恢复原始信号,并且使二进制信号的重建算法更具有实用性。Signal reconstruction algorithm gains great attention in whole field of compressed sensing. In order to solve the problem of signal reconstruction under the condition of sparsity levels being unknowns this paper introduces an energy-based sparsity adaptive matching pursuit （ESAMP） algorithm based on the classical sparsity adaptive matching pursuit （SAMP）algorithm. The proposed algorithm adaptively adjusts step size based on the ratio of measurement vector＇ s energy and the reconstruction signal＇ s energy and thus determines the reasonable initial value of step size. It further corrects the reconstruction algorithm of binary signal, improves the reconstruction accuracy of binary signal, and realises its complete reconstruction. Simulation results show that this algorithm can guarantee higher reconstruction accuracy while speeding up the reconstruction under same condition, and restores the original signals with better comprehensive performance, thus makes the reconstruction algorithm of binary signal becomes more practical.

关 键 词：重建算法 压缩感知 稀疏度 基于能量 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]