Research On DOA Estimation Algorithm Of Passive Coherent Signal

To achieve the purpose of direction finding,passive radar direction system,which does not emit electromagnetic waves itself,receives the signal from the far-field space and processes it to capture the direction of the target signal.Direction of arrival(DOA)estimation,whose main purpose is to estimate the parameters of airspace or the position of signal source,is one of the important directions of array signal processing research,and is widely used in biomedical,earthquake,sonar,radar communications,radio astronomy and other fields.In practical applications,due to the existence of multipath propagation and man-made interference,coherent signals often exist in the incident signal source.However,the classical DOA estimation algorithm can not effectively confirm the source of the coherent signal.To solve this problem,the domestic and foreign scholars have proposed a number of solutions,but there are many limitations in the actual scene.Therefore,it becomes a new research direction to seek high array utilization rate,low computational complexity and high-order decorrelation algorithm.Under the background of passive direction finding,several improved algorithms for DOA estimation are introduced by studying the principle of the existing classical de-coherent preprocessing algorithm in this paper.This thesis is focused on DOA estimation of coherent signals.Firstly,the mathematical models of coherent signals and independent signals are established.The reasons why the traditional DOA estimation algorithm can not be used for coherent source are analyzed.Secondly,the principles of current common de-correlation DOA estimation algorithms are studied,including space smoothing algorithm,vector singular value algorithm,matrix decomposition algorithm,Toeplitz matrix reconstruction algorithm,modified MUSIC algorithms.The advantages and disadvantages of these algorithms are analyzed by computer simulation.Based on the spatial smoothing algorithm,two improved de-correlation algorithms are given.One is the generalized covariance differencing algorithm,which can eliminate the spatial non-uniform noise,adapt to the more general noise field and low noise ratio environment,and reduce the computational complexity.Another is the spatial smoothing algorithm of propagator.This algorithm constructs the spectral function through linear decomposition operation with low computational complexity and high accuracy of direction finding.Finally,an improved algorithm based on the fourth-order cumulant MUSIC algorithm is studied.The MFOC spatial smoothing algorithm is emphatically introduced.This algorithm is analyzed from the perspective of preprocessing coherent sources,and has the advantages of angle estimation of coherent signal sources such as array expansion performance,no data redundancy and suppression of color noise interference.The MFOC spatial smoothing algorithm is especially suitable for the situation where the antenna dish is limited in volume,and the direction of the multi-coherent target signal can be accurately and quickly estimated without losing the antenna aperture.