Enhanced Near-Field Coherent Source Localization via Alternating Oblique Projection and Polynomial Root Finding

Near field (NF) source localization plays an essential role in the field of array signal processing. However, for coherent sources, conventional SSP and Topelitz matrix reconstruction algorithms cannot be applied directly, because of the nonlinear phase variations in near-field source. Although probability model-based methods like enhanced maximum likelihood method (EML) can deal with coherent near-field sources, they are fundamentally constrained by computational complexity or resolution performance. To address these issues, this paper proposes an enhanced method for near-field coherent source localization using a two-step alternating oblique projection (AOP) and polynomial root-finding, comprising a coarse localization stage followed by a fine estimation. Firstly, the coarse estimates of DOA and range parameters are derived by applying the anti-diagonal components of the covariance matrix through root-MUSIC and root-Capon, respectively, followed by an iterative refinement process that alternately applies the AOP method and root-finding to achieve fine estimation. The performance of the proposed algorithm is validated through simulations. Results demonstrate that the algorithm achieves accurate localization of NF coherent sources with reduced computational complexity and improved resolution. The accuracy of the proposed method is also validated by a filed experiment.