An Enhanced RAIM Method for Satellite-Based Positioning Using Track Constraint

Integrity of global navigation satellite system (GNSS) positioning is one major concern for the GNSS-based railway train control systems due to critical safety requirements. In this paper, we propose a map-enhanced receiver autonomous integrity monitoring (RAIM) method for satellite-based positioning of railway trains, which could benefit from the autonomy of fault detection and exclusion (FDE) without relying on additional sensors in a location determination system (LDS). With the proposed RAIM solution, the spatial constraint from the fixed railway tracks is employed to predict the measurements from invisible satellites, which makes it possible of performing FDE even when insufficient satellites are visible to enable the conventional RAIM operation. We demonstrate that different faults in raw GNSS measurements can be identified and isolated by extracting immediate information from the state estimator. With a virtual satellite pseudorange generation mechanism, two-stage FDE architecture is proposed. A pseudorange consistency check logic in the measurement domain is adopted at a local stage. Furthermore, a filter-innovation-based RAIM strategy is utilized to realize a global test in the positioning domain under a cubature Kalman filter (CKF)-based sensor data fusion framework. The proposed method is verified by the field experiment and simulations and the results show the improvement to the availability and safety of GNSS-based positioning.