A novel near-field source localization algorithm without pairing parameters

This paper presents a localization algorithm for near-field narrowband sources in the additive white Gaussian noise environment Firstly, by use of three second-order statistics (SOS) matrices obtained from array outputs, a parallel factor (PARAFAC) analysis model is constructed in the SOS domain; secondly, the uniqueness of the model's low-rank decomposition is analyzed; thirdly, the range, the frequency, and the direction-of-arrival (DOA) of each source are estimated from the decomposed matrices of the model. The proposed method, which does not really require pairing parameters or searching spectral peaks, can be implemented efficiently. Finally, the simulation results are presented to validate the performance of the proposed method.