一种构造压缩感知测量矩阵的新方法  

A new method to construct measurement matrix based on compressed sensing

作　　者：刘晓静[1] 唐加山[2] 

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003 [2]南京邮电大学理学院,江苏南京210023

出　　处：《微型机与应用》2014年第4期74-76,共3页Microcomputer & Its Applications

摘　　要：压缩感知理论是近年来针对稀疏信号提出的一种新的信号处理理论。该理论的主要创新之处在于对信号采样和压缩是同时进行的。测量矩阵是实现该创新点的关键步骤之一,其性能直接关系着信号能不能精确重构。利用行列式非零的对角矩阵的正交性,结合正交基线性表示理论,提出了一种新的更简单的测量矩阵的构造方法。通过实验仿真,验证了新矩阵具有较好的性能。Compressed sensing theory which is proposed recently for sparse signal is a new theory used in signal processing area. It＇s main innovation is to sample and compress signal at the same time. Measurement matrix is one significant step to implement the innovation, and it＇s property has a direct effect on the result of signal reconstruction. This essay proposed a new simple method to construct the measurement matrix , utilizing the orthogonality of diagonal matrix with its determinant of non-zero value, and linear representation theory of orthogonal basis. Simulation result proves that the new matrix has a good performance.

关 键 词：压缩感知 测量矩阵 对角阵 正交基线性表示 

分 类 号：TN911.7[电子电信—通信与信息系统]