Research On DOA Estimation Based On Atomic Norm Theory

Direction of arrival(DOA)estimation plays an essential role in the field of array signal processing.The principle of DOA estimation is essentially to estimate the incoming wave direction of the target source according to the phase difference between the observed signal and the received array.However,in some practical application scenarios,such as different signal models,different noise models and different array models,the performance of the traditional subspace class method will be significantly reduced.In order to overcome the disadvantages,several direction finding methods based on atomic norm theory are proposed in this thesis,which improves the performance of DOA estimation in practical applications.The main work content and innovation of this thesis are as follows:(1)A direction finding algorithm based on atomic norm theory is proposed to estimate the DOAs of coherent signal.To solve the problem that the previous DOA estimation algorithms based on the coprime array will fail when the phase difference of coherent signal meets certain conditions,an atomic norm minimization(ANM)problem is constructed using coprime array interpolation technology,then a low rank Hermitian Toeplitz matrix is recovered,which is not affected by coherent signal phase difference characteristics.Therefore,the DOA estimation based on coprime array for coherent signals can be achieved.(2)A DOA estimation algorithm based on atomic norm theory is proposed to separate circular signals,nonstrictly non-circular signals and strictly noncircular signals.To solve the problem that previous coprime DOA estimation algorithms will fail when the three types of signals coexit,the array interpolation technology is utilized to construct two ANM problems,which are solved to recover two low rank Hermitian Toeplitz matrices.Then the two matrices can be used to construct the space spectrum functions.Therefore,the effective separation of circular signals,non-strictly non-circular rate signals and strictly noncircular rate signals under mixed signals can be achieved.(3)A direction finding algorithm based on atomic norm theory is proposed to realize the direction of departure(DOD)and DOA joint estimation algorithm based on multiple input and multiple output(MIMO)radar for coherent targets.To solve the problem that the traditional parameter estimation methods based on cprime bistatic MIMO radar will fail for coherent targets,a dual-array interpolation technique is designed to construct a Hankel matrix,and the relevant information of the Hankel matrix is used to construct a block Hankel matrix,and then the block Hankel matrix is reconstructed based on the atomic norm theory.The recovered matrix is not affected by the coherence and phase difference between the signals.Therefore,the joint DOD and DOA estimation based on sparse bistatic MIMO radar for coherent targets can be achieved.