Integrity Analysis for Greedy Search Based Fault Exclusion with a Large Number of Faults

Abstract

Detection and exclusion can be achieved using Solution Separation (SS), for example, in civilian aviation applications using Advanced Receiver Autonomous Integrity Monitoring (ARAIM). Global Navigation Satellite Systems (GNSS) fault modes, which are well defined with predictable probabilities of occurrence, can be efficiently and exhaustively accounted for by grouping. However, other navigation applications emerge that require multi-sensor fusion in varying operational environments. Both the total number of measurements and the number of potentially faulty measurements can increase significantly. For such applications, an exhaustive account of all modes can be challenging and can become computationally intractable. These challenges can be addressed using greedy-search-based fault exclusion. However, methods are lacking to quantify the navigation integrity performance of such approaches. In this research, we perform a theoretical analysis of the integrity of greedy-search-based fault exclusion approach. The methods to calculate the minimal detectable bias and the worst case fault direction are discussed.