Bounding Fault Probabilities for Advanced RAIM
Résumé
Advanced Receiver Autonomous Integrity Monitoring (ARAIM) is under development to support both lateral and vertical guidance. A key component of ARAIM is the Integrity Support Message (ISM) which contains validated input parameters including the probabilities of satellite and constellation failure. A service commitment for the former is provided in the Global Positioning System (GPS) performance standards whilst further commitments from GLONASS, Galileo and Beidou are expected to follow. However, Air Navigation Service Providers (ANSPs) may demand some flexibility in setting the ISM. This paper presents a procedure for setting ISM parameters employing a means to determine the a priori probabilities using Bayesian Inference of a Poisson process model given in the literature. This method and a novel way to accumulate the service history are used to analyze the effects of satellite block and age. Results show that whilst the type and age of satellite are clearly drivers in the likelihood of fault, such effects may be accounted for without significant impact on the broadcast values. The paper also addresses the probability of unscheduled satellite outages using the new method. Finally, the paper gives a high-level scheme of how the validity of the ISM parameters may be assured.