Two-dimensional Doa Estimation Algorithm Based On Distributed Source Model With L-shaped Array

Direction of arrival(DOA)is an important part of array signal processing,which can be applied to wireless communication,radar,sonar and many other fields,and has achieved rich research results.However,it is found that the existing results or algorithms are based on the ideal point source model,while the actual wireless signals are often transmitted along multiple paths in the transmission process.In this context,the distributed source model is more compatible with the actual transmission channel environment.At present,compared with the DOA estimation methods under the point source model,the DOA estimation results under the distributed source model are not enough,although there have been some related methods,but they are all based on the ideal array model.so far,the DOA estimation methods based on the distributed source model under the array model error have rarely been reported.It is of great research significance and practical value to study the calibration of array errors and the corresponding DOA estimation methods under the distributed source model in depth.To this end,this paper investigates the DOA estimation method based on single calibration source and Propagator Method(PM)algorithm in the presence of amplitude phase error and coherently/ incoherently distributed sources model,in order to break through the problems of existing distributed source methodsFirstly,the representative DOA estimation algorithms for coherently and incoherently distributed sources are analyzed in depth,and it is found through simulation experiments that the existing methods can obtain high-resolution estimation results only in the ideal array model,however,the performance of the algorithms seriously degrades or even fails when amplitude phase errors exist.The complexity analysis of the DOA estimation algorithm for incoherently distributed sources shows that most of the existing methods perform spectral peak search or iterative operations,which is difficult to apply in the practical environment with high complexity.Secondly,a single calibration source-based gain-phase errors calibration and a two-dimensional DOA estimation method based on PM under the coherently distributed source model are proposed.The method first uses the difference and eigenvalue decomposition of the array covariance matrix of two adjacent sampling periods to complete the estimation of the array gain-phase errors parameters,and then completes the estimation of the distribution source center angle based on the PM algorithm after the compensation of the array gain-phase errors.The simulation experiments confirm that the proposed algorithm can effectively suppress the gain-phase errors noise and provide improved DOA estimation performance with low algorithm complexity.Finally,a new DOA estimation algorithm based on the joint PM and generalized ESPRIT under the incoherently distributed sources model is proposed.The algorithm first approximates the two-dimensional incoherently distributed sources model as a point source model,obtains the initial DOA estimation using the PM algorithm,and then implements the two-dimensional DOA and extended angle estimation of the incoherently distributed sources based on the initial DOA estimation in combination with the generalized ESPRIT algorithm.The effectiveness of the proposed algorithm is verified by simulation experiments,while the algorithm has a lower complexity compared with existing methods.