A NOVEL ANGULAR ESTIMATION METHOD IN THE PRESENCE OF NONUNIFORM NOISE

A novel algorithm for direction-of-arrival (DOA) estimation in nonuniform sensor noise is developed. The diagonal nonuniform sensor noise covariance matrix is estimated by an iterative procedure, which only requires a few iterations. Using the generalized eigendecomposition of two matrices and the least squares, the noise subspace is refined and the noise covariance matrix is estimated iteratively. Since there is no need for knowledge of true DOAs when estimating the noise covariance matrix, our method is superior to most existing approaches. For the proposed noise covariance estimator, we also derive the asymptotic variance of one iteration. Numerical simulations are carried out to demonstrate the advantages of the proposed algorithm over existing state-of-the-art methods.