Direct Position Determination of Non-Circular Sources Based on a Doppler-Extended Aperture With a Moving Coprime Array

Direct position determination (DPD) is a new promising technique in wireless location. Compared with conventional two-step localization methods, DPD achieves a higher accuracy by directly estimating the source location without computing the intermediate parameters. However, all the existing angle-based DPD algorithms for non-circular sources use uniform linear arrays, which lead to low degrees of freedom and poor estimation precision, and do not make use of the Doppler characteristics of the moving station. To improve the DPD performance, this paper proposes a novel DPD algorithm for non-circular sources based on a Doppler-extended aperture with a moving coprime array. First, the coprime array is introduced to angle-based DPD, and an extended array model is established by exploiting the Doppler characteristics of the moving array. The characteristics of non-circular sources are used to further improve the positioning accuracy. Finally, merging the subspace data for each measuring position, the target positions can be obtained. The Cramer-Rao lower bound of the DPD algorithm is presented for non-circular sources in a coprime array combined with Doppler-extended aperture. Performance analysis and the simulation experiments show that compared with the conventional two-step localization method and DPD based on a uniform array, including subspace data fusion and weighted subspace fitting algorithms, the proposed algorithm effectively improves the location accuracy at the expense of a slight complexity increase and can determine the location of multiple targets in underdetermined conditions.