Research On DOA Estimation Algorithm For Coprime Array

Direction of Arrival(DOA)estimation is one of the hot research directions in the field of array signal processing,and has broad application prospects in multiple fields such as radar,5G communication,sonar,and biomedicine.The DOA estimation algorithms based on uniform array require that the spacing between array elements does not exceed the half wavelength of the incident signal,which limits the physical aperture of the array and can’t perform direction finding on signals exceeding the number of array elements.The proposal of the coprime array overcomes this difficulty by having a larger array aperture with the same number of elements,improving the degree of freedom of the array,thereby achieving underdetermined DOA estimation of signals and ensuring good estimation performance.At present,there are still some shortcomings in the research of coprime arrays and their related algorithms.For example,the original structure of coprime arrays cannot significantly improve the degree of freedom of the array,and it can further improve the DOA estimation performance of existing algorithms.Some algorithms are difficult to implement in practical engineering and need to reduce computational complexity.In summary,this article mainly focuses on the study of coprime array including the one-dimensional and two-dimensional DOA estimation algorithms,and proposes the solutions for the problems in the existing algorithms.The main work is as follows:1.In terms of one-dimensional DOA estimation,this paper proposes two DOA estimation algorithms based on the matrix reconstruction of coprime array to improve DOA estimation performance and reduce computational complexity of the existing algorithms for the coprime array.The two proposed algorithms use a coprime linear array to construct a virtual uniform linear array,and use the covariance matrix of the virtual array for matrix reconstruction instead of spatial smoothing.The reconstructed Toeplitz matrix is used as the DOA estimation matrix to complete one-dimensional underdetermined DOA estimation of the signal.Among them,the first method is combined with the estimation of signal parameters via rotational invariance technique(ESPRIT)to perform eigenvalue decomposition on the covariance matrix and rotation invariant matrix,achieving one-dimensional underdetermined DOA estimation of the signal.The second method is combined with the propagator method(PM),only requires eigenvalue decomposition once,further reducing computational complexity.Simulation experiments verify that the two proposed algorithms are effective.2.In terms of two-dimensional DOA estimation,in view of the high computational complexity of the existing 2-D DOA estimation algorithms,this paper firstly proposes an improved algorithm,which uses ESPRIT algorithm to replace the process of sparse reconstruction in the existing algorithm,and greatly reduces the computational complexity on the premise of ensuring the performance of DOA estimation.In addition,this paper also proposes a parallel expanded coprime linear array and its corresponding two-dimensional DOA estimation algorithm.Using the parallel expanded coprime array,a virtual parallel uniform linear array can be constructed.Compared with the traditional parallel coprime linear array,it can make full use of the array element information,and can use the autocovariance matrix and cross covariance matrix of the virtual array to construct Toeplitz matrix respectively,Then,the augmented matrix is constructed and its eigenvalue decomposition is performed,which transforms the two-dimensional DOA estimation into two one-dimensional problems,improves the performance of DOA estimation and reduce the computational complexity..Simulation experiments verify that the two proposed algorithms are effective.