稀疏信号重构的阈值化迭代检测估计

发布时间：2019-06-06 04:22

【摘要】：研究压缩传感(Compressed Sensing,CS)的稀疏信号重构算法,该文提出一种新的算法框架阈值化迭代检测估计(Iterative Detection Estimation with Thresholding,IDET)。算法框架包括两个方面:选择单阶段阈值化(One-Stage Thresholding,OST)算法的迭代步作为支持集检测的参考;根据稀疏信号的特征设计支持集检测方法。同时,提出该算法框架的实现算法,实现算法先检测由迭代硬阈值化(Iterative Hard Thresholding,IHT)迭代步得到一个支持集,然后通过求解支持集上的最小二乘问题来估计待重构的稀疏信号,迭代上述两个步骤直至满足条件停止。IDET算法的关键在于支持集检测,该文提出3种适用于快速衰减信号的支持集检测方法。实验结果表明,IDET稀疏重构性能优于IHT的其他加速算法。
[Abstract]:In this paper, a sparse signal reconstruction algorithm for compressed sensing (Compressed Sensing,CS) is studied. A new algorithm framework for threshold iterative detection and estimation of (Iterative Detection Estimation with Thresholding,IDET is proposed in this paper. The algorithm framework includes two aspects: the iterative step of single-stage threshold (One-Stage Thresholding,OST) algorithm is selected as the reference for support set detection, and the support set detection method is designed according to the characteristics of sparse signal. At the same time, the implementation algorithm of the algorithm framework is proposed. Firstly, a support set is obtained by detecting the iterative step of iterative hard threshold (Iterative Hard Thresholding,IHT), and then the sparse signal to be reconstructed is estimated by solving the least square problem on the support set. The key to IDET algorithm is to support set detection. In this paper, three support set detection methods for fast attenuation signals are proposed. The experimental results show that the sparse reconstruction performance of IDET is better than that of other accelerated algorithms of IHT.
【作者单位】：江苏大学计算机科学与通信工程学院;中山大学信息科学与技术学院;
【基金】：国家自然科学基金(61375080,61170126) 江苏省科技支撑计划项目(BE2013696) 江苏大学高级人才科研基金(12JDG050)资助课题
【分类号】：TN911.23

【相似文献】