Mixed Far-Field and Near-Field Source Localization Using a Linear Tripole Array

Abstract

This letter presents a novel algorithm for mixed far-field (FF) and near-field (NF) source localization using a linear tripole array. Firstly, several special fourth-order cumulants are constructed, then the shift invariant properties in the cumulant domain are utilized to estimate the parameters of each source at two tripoles with spacing exceeds a half wavelength. On this basis, a matrix equation is derived to classify the sources types and estimate the parameters of mixed sources at the phase reference point. Finally, the obtained locate estimates can be used to disambiguate the cyclically ambiguous estimates to produce the high-accuracy parameters estimation. Compared to the existing methods, the proposed one doesn’t need to know each sensor’s position and not restrict the array configuration to be uniform linear array (ULA). Moreover, it estimates both the spatial and polarization information of mixed sources and provides improved accuracy without any spectral search, as well as realizes a more reasonable classification of the signal types. Numerical simulation results show the effectiveness of the proposed method.