基于压缩感知的信号频谱测量方法.pdfVIP

基于压缩感知的信号频谱测量方法 摘 要 传统的信号频谱测量方法是基于香农-奈奎斯特定律进行采样，并根 信息 本身的特性降低它的冗余度从而实现对它的压缩，经过编解码之后再对获得的 信号进行频谱测量。近年来出现的压缩感知理论(Compressed Sensing,CS)突破了 传统香农-奈奎斯特定律的限制，它是通过非自适应的线性投影来使信号保证原 始结构，并将采样和压缩两个过程合二为一，尽量从更少的数 中获得更完全 的信号信息，使频谱测量更为简单高效。 本文介绍了压缩感知理论的基本原理框架及信号稀疏变换、观测矩阵设计 和重构算法这三个关键环节的最新进展，讨论了目前研究中存在的重难点问 题。并在Matlab仿真平台上，分别在傅里叶变换(The Fast Fourier Transform,FFT) 及离散小波变换(Discrete Wa elet Transform,DWT) 的前提下，采用正交匹配追踪 算法(Orthogonal Matching Pursuit,OMP)实现了对一维信号和二维图像信号的基 于压缩感知的高概率重构。通过比较重构结果与原始信号，发现该方法只需要 远小于传统采样方式的采样率，就能得到原始信号包含有用信息，再利用压缩 感知重构算法就能实现高概率且精确地重构。 关键词 ：压缩感知 ；正 ； ；测 The Signal Spectrum Measurements Based on Compressed Sensing Abstract The traditional technology of the signal spectrum is based on the Shannon- Nyquist sampling, and reduces its redundancy feature based on the information itself, reaching the goal of compression, and measures the signal spctrum after re-encoding and decoding.In recent years, Compressed Sensing(CS) which does not ha e to obey Nyquist sampling theorem comes out as a new sampling theory, and it can keep the original structure of signals by attaining the non-adapti e linear proj ections. It make the sampling and compressing in combination. So, Compressed Sensing can gather more data and more information directly from less data and make the spectrum measurement more simple and efficient. This paper introduces the basic principles of the framework of the Compressed Sensing theory and the latest de elopments of three keys including signal sparse representation, design of measurement matrix and reconstruction algorithm.Then this paper also discusses the existing difficult problems. Based on the Fast Fourier Transfo