This paper describes the design and preliminary evaluation of a new 'Online' ground monitor for advanced receiver autonomous integrity monitoring (ARAIM). ARAIM is intended for vertical guidance of aircraft down to 200 feet altitude. In ARAIM, fault detection is autonomously performed at the airborne receiver using dual-frequency, multi-constellation GNSS. The ARAIM ground monitor aims at validating the assertions made at the airborne receiver on 'integrity parameters', which include, for example, the ranging error variances and prior fault probabilities associated with satellite orbit and clock ephemeris errors. To determine these integrity parameters, two candidate architectures, Offline and Online, are under consideration. Both architectures assume the constellation service providers (CSP) satisfy their service performance commitments. But, in addition, Online ARAIM provides a set of precise 'overlay' orbit and clock ephemeris parameter predictions to replace the CSP navigation message. And, the Online monitor can broadcast hourly updates on the integrity parameters. It follows that probability bounds on integrity parameters can be tightened, which ultimately reduces availability risk as compared to the Offline architecture. On the other hand, Online ARAIM increases 'connectivity risk', as it requires monitor updates to be transmitted from the ground segment to the aircraft. This paper presents an Online monitor concept, which establishes a clear relationship between the monitor's detection test statistic, and the integrity parameters broadcast to the aircraft. The test statistic is derived independently from the Online overlay ephemeris, using code and carrier phase measurements collected at a few worldwide reference stations. Preliminary evaluation of the test statistic distribution is carried out using truth data from the International GNSS Service (IGS), and using covariance analysis. Results suggest that SV position and clock estimation, which is needed to establish the monitor test statistic, is achievable at the decimeter level using a sparse network of ground reference stations.
展开▼
