Indexed by:Journal Article

Date of Publication:2017-01-01

Journal:DIGITAL SIGNAL PROCESSING

Included Journals:EI、SCIE

Volume:60

Page Number:242-251

ISSN:1051-2004

Key Words:Correntropy; Direction of arrival; Symmetric alpha stable processes; MUSIC algorithm

Abstract:By introducing correntropy as the robust statistics, a novel direction of arrival estimator for a-stable noise is proposed. In this method, the signal subspace is estimated by solving the correntropy based optimization problem under the maximum correntropy criterion. An optimal step size based iterative algorithm is developed and the convergence of it is proved. Comprehensive simulation results demonstrate that the proposed method is superior to several existing algorithms in terms of the probability of resolution and the estimation accuracy, especially in the highly impulsive noise environments. (C) 2016 Elsevier Inc. All rights reserved.