Previous Abstract Return to Session C4 Next Abstract

Session C4: Navigation Using Environmental Features

Context-Adaptive GNSS/INS Estimation Strategies Based on Prior Statistical Multipath Knowledge
Maximilian Von Arnim, Damien Vivet, and Yoko Watanabe, Fédération ENAC ISAE-SUPAERO ONERA, Université de Toulouse
Location: Grand Ballroom IJ
Date/Time: Wednesday, Apr. 30, 2:58 p.m.

Modern mobile applications demand accurate and ubiquitous positioning, requiring approaches that explicitly account for environmental context. A major challenge in Global Navigation Satellite System (GNSS) positioning is multipath interference, which remains difficult to characterize and correct. This issue is particularly pronounced in low-cost GNSS receivers limited to pseudorange observables, which are highly susceptible to multipath errors. In this work, we present a novel statistical multipath model based on GNSS data collected in and around Toulouse, France, across four distinct environmental contexts: urban canyons, open sky, tree-covered areas, and general urban settings. These statistical models were derived using high-precision reference receivers. Leveraging this model, we propose a context-aware adaptation of standard approaches in GNSS positioning such as weighted least squares and extended Kalman filter estimators. Additionally, our method integrates tightly coupled Inertial Navigation System (INS) data and Doppler shift measurements, while employing robust outlier rejection techniques. We validate our approach using a second independent dataset collected in Carcassonne, France, demonstrating its effectiveness in improving positioning accuracy. Finally, we discuss potential extensions towards further context-adaptive GNSS estimators, aiming to enhance real-world applicability in diverse environments.
Index Terms—Environmental Context, Adaptive Estimation, Multipath Model

Previous Abstract Return to Session C4 Next Abstract