压缩感知重构匹配类算法分析

A survey of reconstruction algorithms based on matching in compressive sensing

压缩感知理论是一种利用信号的稀疏性或可压缩性而把采样与压缩融为一体的新理论体系,它成功地克服了传统理论中采样数据量大、资源浪费严重等问题。该理论的研究方向主要包括信号的稀疏表示、测量矩阵的设计和信号的重构算法。其中信号的重构算法是该理论中的关键部分,也是近年来研究的热点。本文主要对匹配追踪类重构算法作了详细介绍,并通过仿真实验结果对这些算法进行了对比和分析。

The compressive sensing theory is a recent proposed theory, which can utilize the sparse and compressive characteristics of signals to combine the sampling and compression processes into only one procedure. It overcomes the shortcomings of large sampling data and significant waste of resource in traditional theories. The research areas in compressive sensing mainly include the sparse representation of signals, the design of measurement matrix, and signal reconstruction algorithms. The reconstruction algorithm is the key component of this new theory and is the focus of recent research. In this paper, the reconstruction algorithms, which use matching and tracking techniques, are described in details. Simulations of these algorithm are also conducted to compare and analyze these algorithms.