A one-step algorithm for mixed far-field and near-field sources localization

Abstract

In this paper, a one-step method is proposed for the classification and estimation of the parameters in the scenario of mixed near-field and far-field stationary sources. By constructing several temporal-spatial cumulant matrices, utilizing the stationary property of sources, and introducing a mechanism for controlling the rank, a parameters estimation matrix is formed. This matrix contains electric angles and signals information. By introducing a method based on the rotational invariant technique and using the estimation signal parameter, the direction-of-arrival of all sources' signals and the range of the near-field sources (NFSs) are estimated. Also, a reasonable classification of the signals types is provided. Unlike conventional methods, the proposed method does not require any spectral search stage, so it reduces the computational complexity to a great extent. The proposed algorithm prevents the array aperture loss, has no need to know the number of NFSs or field-field sources, and does not require any extra stage to pair the parameters. Mathematical analysis and simulation results validate the good performance of parameters estimation and sources classification in the proposed method.