Study On Application Of Compressed Sensing In Underwater Echo Processing

Underwater target echo signals are the basis of active sonar detection and recognition.Underwater environment is complex and varied,there are many kinds of noise and interference.Especially with the development of stealth technology and the emergence of a variety of small targets,the echo signal of target is getting weaker.The weak echo signal detecting and processing problems are becoming a hot and difficult issue in the field of underwater signal processing.In addition,in order to improve the system's detection accuracy,increase the anti-interference ability and improve the target detection probability,we need to increase the signal bandwidth.However,the increase of bandwidth makes the data volume increase sharply,which brings great burden to signal acquisition,storage,transmission and processing.Under this kind of background,this thesis studies the application of compressed sensing and tries to solve the problems such as weak signal detection and oversampling data in underwater echo signal processing.The main work is as follows:Firstly,the research background and significance of this paper are expounded,and the theory of compressed sensing and its application in underwater signal processing are reviewed,and the research ideas and structure of this paper are given.Secondly,based on the theory of compressed sensing,several common sparse bases,measurement matrices,matching tracking reconstruction algorithms and corresponding improved algorithms are studied.The measurement standard of signal reconstruction effect is given.Based on the basic theory of underwater echo signal,the simulation of underwater echo signal is realized by using the bright point model.Based on this,the influence of different sparse matrix,different measurement matrix and different reconstruction algorithms on the results of underwater echo signal processing is compared through simulation experiments,and a framework of compressed sensing processing based on discrete cosine sparse basis,Gaussian random matrix and segmented orthogonal matching pursuit algorithm is proposed.Thirdly,the formation principle and structural characteristics of the underwater echo signal are fully analyzed.The relationship between the echo signal and the incident signal,the block characteristic of the echo signal is studied.The sparse decomposition and reconstructing process is introduced into the sparse decompositionof the incident signal and the block sparse characteristic as a prior information.A compressed sensing processing method incorporating the a prior information is proposed and further applied to the underwater echo signal processing.The improvement of signal-to-noise ratio(SNR),matching degree and relative error are used to measure the processing effect of this method,which shows the advantage of this method in improving signal-to-noise ratio and reducing data volume.Finally,the main work and innovation of this thesis are summarized,and the future research should be carried out.