New concept of adaptive beamforming for moving sources and impulse noise environment

Abstract

A minimax robust two-dimensional beamforming is developed for complex-valued observations contaminated by impulse random errors having a unknown heavy-tailed distribution. Estimates of directions of arrival (DOA) are defined by minimizing a nonquadratic residual function derived from the Huber's minimax robust estimation theory. A tracking ability of the estimates is assured by using the deterministic nonparametric model of source movement and a sliding window of observations. The proposed new beamformer has a two-dimensional power function. Maximum peaks of this power function are used for source separation and estimation of DOAs and their first derivatives. Simulation demonstrates that the new beamforming is able to solve tight sources, as well as to give high accuracy estimates for rapidly moving sources. Besides, the new estimates show the strong resistance to heavy-tailed distribution noises.