Near-Field Source Localization Using the SPICE Method for Limited Snapshots

Due to the highly nonlinear phase of received near-field signals, most near-field source localization methods are developed using approximate signal models. However, estimation methods based on these approximate signal models suffer from model mismatch and performance degradation, especially in scenarios with limited snapshots. In this paper, we propose a near-field source localization method based on an exact signal model using the Sparse Iterative Covariance-based Estimation (SPICE) algorithm. We construct a sparse dictionary based on the near-field source signal model received by the sensor array and analyze the analytical forms of the optimal solutions under both limited and sufficient snapshot scenarios. Through iterative methods, we estimate the sparse representation of the signals, thereby achieving source localization and separation. Numerical simulations are provided to demonstrate the effectiveness of the proposed method.