An Evaluation of the Advanced RAIM Threat Model

Abstract

Protection Levels in RAIM and Advanced RAIM are designed to protect against worst-case biases in the satellite or combination of satellites that could be faulted. When examining the curves of probability of missed detection as a function of magnitude, this may seem to be an excessively conservative requirement, given that the fault biases are unknown (and therefore random to a certain extent). In this note, we consider relaxations of this worst-case requirement by assuming that the fault biases have known distribution. We find that even optimistic distributions of the fault biases are unlikely to yield significant benefits in typical ARAIM scenarios.