Modified spectral function based DOA estimation in colored noise

With the wide application of large-scale antenna arrays in the field of radar, sonar and wireless communication, direction of arrival (DOA) estimation based on the general asymptotic theory (GAT) has received more and more attentions recently, which can effectively conquer the inconsistent estimation problem of sampling covariance matrix (SCM) in the scenarios that the number of sensor M is large and of the same order of magnitude as the number of snapshots N. From the GAT perspective, a new DOA estimation method (termed as SEMDT-M-MUSIC) in colored noise is investigated in this paper, which mainly comprises three steps, the first step is to obtain an improved SCM through covariance matrix shrinkage estimation, the second step is to eliminate the influence of the colored noise via the matrix difference technology, and the last step is to enhance the DOA estimation via the modified spectral function, where the phase transformation result of the spike covariance matrix is employed. Numerical simulations validate the effectiveness of the proposed method.