The Estimation Of Sources Enumeration Under Color Noise And Small Snapshots

In array signal processing,direction-of-arrival(DOA)estimation as an essential research orientation,which has a very broad application prospect in many fields such as wireless communication,radar,sonar,navigation,biomedical signalmeasurenciat,etc.In the past 30 years,DOA estimation algorithms have made great progress,and the requirements for accuracy and resolution are getting higher and higher.However,most DOA estimation algorithms with high precision and super resolution are required to obtain the amount of signal sources.Therefore,the sources enumeration is of great significance to the subsequent DOA estimation.At present,many signal source number estimation algorithms are based on Gaussian white noise and the sampling number of snapshot is large enough.However,in practical engineering,the noise background is mainly spatial color noise,and the number of samples can not meet the large demand,which makes the performance of these signal source number estimation algorithms decline sharply or become invalid.Papers on spatial colored noise background as well as small snapshot number of source number estimation problem,based on the DMR whitening filter and enhance the estimation method of covariance matrix,the method of DMR whitening filter used to solve the problem of space color noise,covariance matrix and optimal convex linear combination of unit matrix to enhancement of covariance matrix,The improved covariance matrix is used to estimate the number of signal sources according to the information theory criterion.The main work substances of the thesis contain the following points:1.This paper introduces the research background of source number estimation and the research status of domestic and foreign scholars in the field of this algorithm in detail,and puts forward the research objectives according to the status quo.Then,taking MUSIC algorithm as an example,the influence and significance of source number on DOA spatial spectrum estimation are simulated and analyzed.2.Several classical signal source number estimation algorithms under spatial color noise and two important theoretical pillars DMR technology and enhanced covariance matrix technology are introduced.Classic algorithms including information theory algorithm based on diagonal loading technique,Gerschgorin method,and based on gerschgorin radius and the rule of information theory,through the simulation experiment data of information theory criterion respectively under white gaussian noise and colored spatial noise estimation of difference analysis,and comparative analysis of each algorithm under different influencing factors estimation performance,etc.The theoretical knowledge of these two technologies is analyzed,and their functions in array signal processing are described.3.This paper focuses on the design of the overall flow of the algorithm.In order to solve the problem of divergence of noise eigenvalues under spatial color noise and degradation of the performance of most source estimation algorithms under small snapshot number,a new source estimation algorithm based on DMR whitening filter and enhanced covariance matrix is proposed in this paper.The DMR whitening filter is constructed by combining the pure noise covariance matrix with DMR algorithm,and then the DMR whitening filter is used to whiten the estimated covariance matrix containing signal and noise.Then the optimal convex linear combination of the whitened covariance matrix and the identity matrix is used to improve the accuracy of the estimated covariance matrix.Finally,the improved covariance matrix was decomposed by eigen value decomposition(EVD)to obtain more accurate eigen value,and the number of signal sources was determined by combining AIC and MDL criteria.4.Simulation results verify the effectiveness and superiority of the proposed algorithm.Simulation experiments verify that the DMR whitening filter can solve the problem of noise feature divergence caused by spatial color noise,and verify that the proposed algorithm can also get a more accurate covariance matrix under a small snapshot number.Finally,compared with the other four algorithms in the estimation performance,it is proved that the algorithm in this paper is better in the estimation of the number of sources.