Variational Inference-Based Multi-Target Localization Under Impulsive Noise

This paper presents a novel method for multi-target counting and localization based on Variational Bayesian Inference (VBI), which addresses the challenge of multi-target localization under impulsive noise conditions. Specifically, this study first formulates a VBI-based target counting and localization model tailored for sparse outliers from impulsive noise. An adaptive Bayesian network framework is then developed to dynamically determine the number of targets and their positions. VBI is then employed to achieve joint estimation of outlier variables, target number, and target positions. Simulation results demonstrate that the proposed approach achieves significant improvements in localization accuracy under impulsive noise conditions when the number of targets is unknown.