摘要
为了解决压缩采样匹配追踪(CoSaMP)算法在观测值固定时重构概率随着稀疏度的增加急剧下降问题,基于最优选择思想和回溯思想设计一种最优选择的压缩采样匹配追踪(OSCoSaMP)算法。在每次迭代过程中,从支撑集中选出最优的支撑,同时采用回溯思想剔除错误原子,分别测试不同稀疏度和不同观测值下的重构概率。仿真结果表明,该算法重构概率与OMP和CoSaMP算法相比有所提升。OSCoSaMP算法在稀疏度50时的重构概率保持在90%以上,当观测值大于70时重构概率在90%以上。OSCoSaMP算法能够有效提高一维信号的重构概率。
In order to solve the problem of compressed sampling matching pursuit(CoSaMP)algorithm when the observation value is fixed,the reconstruction probability decreases sharply with the increase of sparsity.Based on the optimal choice and backtracking ideas,an optimally selected compressed sampling matching tracking(OSCoSaMP)algorithm is designed.In each iteration,the optimal support is selected from the support set,and the backtracking idea is used to eliminate errors.Atom,tested the reconstruction probability under different sparsity and different observation values.The simulation results show that the reconstruction probability of the algorithm is improved compared with the OMP and CoSaMP algorithms.The reconstruction probability of the OSCoSaMP algorithm remains above 90%when the sparsity is 50,and the reconstruction probability is 90%when the observed value is greater than 70.The OSCoSaMP algorithm can effectively improve the reconstruction probability of one-dimensional signals.
作者
刘强
司伟建
LIU Qiang;SI Wei-jian(College of Information and Communication Engineering,Harbin Engineering University;Key Laboratory of Advanced Marine Communication and Information Technology,Ministry of Industry and Information Technology,Harbin 150001,China)
出处
《软件导刊》
2021年第9期78-82,共5页
Software Guide
基金
国家自然科学基金项目(61671168,61801143)
中央高校基本科研业务费专项基金项目(3072019CF0801)
黑龙江省自然科学基金项目(LH2020F019)
航空科学基金项目(2019010P6001)。
关键词
压缩感知
重构算法
贪婪匹配追踪
正交匹配追踪
压缩采样匹配追踪
compressed sensing
reconstruction algorithm
greedy matching pursuit
orthogonal matching pursuit
compressed sampling matching pursuit
