SOURCE NUMBER ESTIMATION FOR UNBALANCED ARRAYS USING ROBUST OUTLIER DETECTION IN THE EIGENVALUE DOMAIN

In this paper, we address the problem of estimating the number of sources impinging on an array of sensors in the presence of unknown non-uniform noise. In such a situation, sensor noise levels across the array are spatially inhomogeneous. We first consider the eigenvalues of the array correlation matrix as a set of measured data, and then we treat the eigenvalues corresponding to the sources as outliers. Thus, the robust source detection in array processing is viewed as a problem of outlier detection in the eigenvalue domain. Then, we propose a new source detection method based on multiple test procedures that considers ordered differences of the robust distance estimates. The proposed approach can be used in uniform/non-uniform noise, non-Gaussian noise, and colored noise. Unlike the information theoretic criteria, which depend on the selected model, our technique can be applied for many array processing models, and is therefore favourable in real world applications.