An Alternative Approach for Pseudorange Variance Estimation Under Scintillation Environments Using Markov-Rao-Blackwellized Particle Filtering

Abstract

Ionospheric scintillations, arising from variations in phase/amplitude of radio signals traversing the ionosphere, pose significant challenges to Global Navigation Satellite System (GNSS) positioning, particularly in low-latitude regions. This paper proposes a Rao-Blackwellized Particle Filter (RBPF) integrated with a Markov chain model to comprehensively characterise and mitigate the impact of ionospheric scintillation on GNSS positioning. Unlike traditional methods, the Markov-RBPF framework offers enhanced versatility in assessing scintillation dynamics both spatially and temporally, allowing for precise modelling of scintillation evolution over varying nighttime hours and months of the year. Through simulations, the authors demonstrate the superior performance of the proposed Markov-RBPF compared to conventional Extended Kalman Filters (EKF), with position root-mean-square errors below 2 m in a scenario of strong scintillation events in October 2014. This showcases its robustness and versatility in improving GNSS positioning accuracy amidst challenging ionospheric conditions.