Distributed Pseudolinear Information Filtering Approach With Multisource Measurements for Heterogeneous Sensor Networks

Abstract

This paper presents a distributed estimation approach for heterogeneous sensor networks based on the information form of pseudolinear Kalman filters (PLKF). First, an integrated PLKF is developed to process multisource range, bearing, and Doppler radial-velocity measurements by converting them into a pseudolinear measurement equation, thus completing a unified PLKF framework for heterogeneous sensing modalities. Then, the instrumental variable-based integrated pseudolinear information filtering (IV-IPIF) algorithms are derived for distributed implementation. Finally, a consensus-based distributed IV-IPIF approach is proposed for heterogeneous networks, with estimation-error boundedness rigorously established. Simulations and experiments on unmanned aerial vehicles and unmanned ground vehicles show that the proposed approach achieves consistent and robust estimation under high prior uncertainty, with accuracy close to the posterior Cramér-Rao lower bound. Note to Practitioners—This article studies cooperative target-state estimation in heterogeneous sensor networks using multisource measurements. Robust real-time estimation is crucial for multi-agent situational awareness. Existing sigma-point–based filtering methods are limited in modeling nonlinear measurement functions with sufficient fidelity. The proposed pseudolinear-transformation framework enables efficient and precise updates, with an information-form extension supporting heterogeneous networks. Results show superior robustness under high prior uncertainty, along with improved accuracy and consistency compared to traditional filters. The approach also applies to mobile robot navigation, environmental monitoring, sports analysis, and Internet of Things applications.