Channel Coding for Compressed Sensing Measurement Matrix
DONG Xiao-liang,YANG Liang-long,ZHAO Sheng-mei,ZHENG Bao-yu
DOI: https://doi.org/10.3969/j.issn.1003-0530.2013.07.003
2013-01-01
Journal of Signal Processing
Abstract:Compressed sensing,which is the emergence of a signal processing theory sparse signal and compressible signals in recent years.The measurement matrix is a vital link in the compressed sensing theory,its signal sampling and reconstruction algorithm has an important impact.Although the traditional random measurement matrix for the reconstruction is quite good,but its hardware implementation is difficult and requires a lot of storage space and other defects.While The emergence of the deterministic measurement matrix,makes up for these shortcomings.Using the advantages of the channel coding check matrix,we put forward the way to meet the requirements of the restricted isometry property,through the constructor of the deterministic measurement matrix.We make the standardization of a parity check matrix of the column vector,and extend it to a square linear combination of the permutation matrix column vector,then a deterministic measurement matrix can be created.This method ensure us to produce the measurement matrix easily,after we complete a channel encoded parity check matrix.Numerical results show that,under the same reconstruction algorithm and compression ratio,the performance of this method is close to the random measurement matrix,even improved.The same time,it costs less time with the reconstruction being run once only,which can meet the real-time requirements.The practical application of the compressed sensing algorithm,provides an effective measurement matrix construction method.