Adaptive Robust Navigation for Safety-Critical Railway Applications

GNSS is recognized as a game changer for the next railway signaling solutions since July 2021 by the EU Parliament. However, the satellite-based technology still faces major issues before full introduction. For safety critical functions, the positioning on-board unit needs to show its compliance with very stringent requirements. To ensure position availability and accuracy, solutions will result from a multisensor fusion (GNSS, IMU, odometer) supported by the use of digital maps and some barriers for error mitigation such as Fault Detection and Exclusion techniques. Main constraints for GNSS performance is caused by track surroundings (building, vegetation, bridges or tunnels) but interference also needs to be detected and managed as the train has to travel in public areas, close to highways or other infrastructure where jamming events in particular are now frequently observed. In this paper, based on recent development for railway applications, we outline a concept for a smart adaptation and mitigation of these effects, merging work performed for context detection, interference detection and classification and different weighted least squares. The different contributions will be described as well as the concept proposed.