Fast Direction Of Arrival Estimation Based On Special Array Configurations

As one of the important branches of array signal processing,direction-of-arrival(DOA)estimation technology has a wide range of application background,and it has always been a research hotspot in array signal processing.Since its inception,DOA has experienced the development process from high resolution to super resolution,although the estimated performance is improved increasingly,the super-resolution estimation algorithm are difficult to be transformed from theory to engineering application because their heavy computational complexity,therefore,it has become another hot issue that how to reduce the computational complexity of them.In this paper,based on the uniform linear array(ULA)and the co-prime array,three fast algorithms for reducing the computational complexity are proposed to solve the problem of large computational complexity,which would provide the theory reference for the application of super-resolution estimation algorithms.The main contents of this paper are as follows:First,after intensively studying the computationally efficient subspace algorithm for 2-D DOA estimation with L-shaped array(CESA),based on its azimuth estimation principle,a new fast DOA estimation algorithm based on the ULA is proposed.The new algorithm divides the ULA into two sub-arrays with equal number of elements,then the forward and backward cross-covariance matrices of the sub-arrays are computed,combined which a joint cross-covariance matrix is constructed,and the signal subspace is obtained by conducting several linear operations with the first column vector of the joint cross-covariance matrix.Simulation experiments show that the algorithm can effectively reduce the computational complexity of the covariance matrix computing and the subspace decomposition while ensuring that the estimation accuracy is acceptable.Next,after a further study of the computationally efficient subspace-based method for two-dimensional direction estimation with L-shaped array(CODE),a new subspace fast estimation algorithm based on the ULA is proposed.The algorithm firstly divides the ULA into two sub-arrays and the array manifold matrix is decomposed into two sub-matrices with rotational invariant relation.According to the relationship between the cross-covariance matrices and the manifold matrix,the cross-covariance matrix is also divided into two sub-matrices which satisfy the rotation invariance relation similarly.Finally the sub-matrix is utilized to solve the rotation invariant relation matrix,with which the subspace is found and a polynomial is constructed to solve DOA.The new algorithm improves the estimation accuracy of the CODE algorithm,and reduces the computational complexity of subspace acquisition effectively.Finally,a new fast DOA estimation algorithm is proposed for the co-prime array which is one focus of research at this stage.A co-prime array is regarded as consisting of two ULAs due to its partial uniform linear structure property in the new algorithm,and then the polynomials of ULAs are established according to the root-MUSIC algorithm.Finally,the polynomial of the co-prime array is constructed by the sub-array polynomials and the roots on the unit circle of the co-prime array polynomial are corresponding to the DOAs we wanted.The simulation results demonstrate that the proposed algorithm outperforms the root-MUSIC algorithm and other existing similar algorithms in the accuracy and efficiency.