Robust AOA-Based Source Localization in Correlated Measurement Noise via Nonconvex Sparse Optimization

摘要

This letter considers the problem of robust angle of arrival (AOA) source localization in the presence of correlated noise and outliers based on nonconvex sparse optimization. To maintain the total convexity of the cost function while using the nonconvex penalty, we propose to regularize the correlation matrix weighted cost by generalized minimax concave (GMC) function. Then the alternating direction forward-backward splitting algorithm (ADFBS) is proposed to estimate outliers and source position simultaneously. To counter the bias problem of ADFBS, a new ADFBS based instrumental-variable estimator (IVADFBS) is developed. The IVADFBS is observed to produce nearly unbiased estimates with lower mean squared errors.