Performance analysis of DOA estimation in the threshold region

This paper presents a performance analysis of Maximum Likelihood (ML) Direction-Of-Arrival (DOA) estimation in the threshold region using a sensor array. This region is a range of Signal-to-Noise Ratios (SNR) close to the threshold, where the estimation error due to outliers starts to rise very rapidly as the SNR decreases. Approximate expressions for the probability of outlier and the mean square estimation error are derived for the case of a single signal in white Gaussian noise and a single snapshot. It is verified by simulations that the approximations predict the ML DOA estimation performance with high accuracy also at low SNR where the Cramér-Rao bound is far too optimistic.