Distributed target tracking via UWSNs in the presence of multipath interference

Abstract

This article addresses the practical challenge of robust target tracking in a distributed network of underwater acoustic sensors operating under multipath interference. In underwater environments, multipath effects can cause received signals to interfere at the transducer, leading to the degradation of acoustic echoes. Consequently, this degradation introduces autocorrelated biases into the original measurements, thereby reducing tracking accuracy. To tackle this issue, we adopt a state-augmentation approach combined with Gaussian filtering to develop a novel distributed filter for a class of nonlinear time-varying systems. By augmenting both the target states and multipath-induced biases, the proposed method effectively handles the nonlinearities and interdependencies between state variables and multipath autocorrelation during the estimation process. We refer to the proposed method as DUKF-Mp and provide theoretical analysis to investigate the stability by verifying its stochastic boundedness. Numerical simulations validate the proposed method, showing that DUKF-Mp outperforms existing approaches in tracking accuracy and maintains robustness even under high levels of multipath interference.