The Global Positioning System (GPS) is already being used for certain aviation applications and some safety critical air traffic services will be based on GPS. These air traffic services must achieve allowable levels of safety before they can be accepted. For this to occur, GPS based navigation systems must achieve a defined level of performance specified in terms of accuracy, integrity, continuity and availability. This must be determined by various analysis techniques including failure mode and effects analysis (FMEA) and integrity assessment. Because of the high percentile requirements placed on integrity (as the parameter most directly related safety), it is unfeasible to measure system performance by demonstration (field trial). Realistic simulation informed by some field experience is usually employed. However, the current simulation-based approaches for receiver autonomous integrity monitoring (RAIM) performance assessment have a number of weaknesses including the use of coarse (large) spatial and temporal sampling intervals, loose definitions of error and geometric correlations, a lack of sampling of all geometries and the inability to account for critical points due to uncorrelated factors.
This paper proposes a dynamic sampling method that takes account of these weaknesses, identifying dynamically only the required points for integrity performance assessment. Comprehensive simulations carried out to test the proposed approach for a single point, an area, and a non-precise approach (NPA) flight path to Gatwick airport in the United Kingdom show that the method can be effective in capturing all the points enabling a robust and reliable assessment of system integrity.