On the Performance of a Cascade of Barriers of Multipath Feared Events for High Accuracy High Integrity Train Positioning

Detection of multipath feared events is a fundamental capability of current and future High Accuracy High Integrity positioning systems for rail transport based on Global Navigation Satellite Systems. In this work, a method based on a cascade of barriers is proposed for Signal In Space measurement selection and integrity monitoring. To this aim, a mathematical framework for the analytic assessment of the achievable performance of Fault Detection and Exclusion algorithms based on Solution Separation is defined, accounting for the impact that exclusion of a Signal In Space observable has on the PVT accuracy and integrity. An adaptive logic based on Gaussian Mixtures models is also introduced for the computation of the exclusion thresholds in order to assure a constant False Exclusion probability.