Integrity Monitoring for GNSS Positioning Via Factor Graph Optimization in Urban Canyons

Fault detection and exclusion (FDE) is significant for integrity monitoring of GNSS positioning autonomous systems with navigation requirements. Moreover, the urban canyon scenario introduces additional challenges to existing FDE monitoring, causing missed or false alarms, due significantly increased percentage fault measurements. This paper proposed a sliding window aided based on factor graph optimization (FGO) alleviate these key issues. Different from snapshot-based sequential-based (e.g. Extended Kalman filter) methods where only current two consecutive epochs measurements are considered in process, method this improves measurement redundancy help structure. Meanwhile, inside multiple times which enables reconsideration FGO employs iterations re-linearizations initial guess state estimation FDE. The effectiveness verified through challenging dataset collected canyons Hong Kong using automobile-level low-cost receivers.