GNSS Inter-Constellation Phasing: Validation of the Worst-Case Assumption

Advanced Receiver Autonomous Integrity Monitoring (ARAIM) uses satellite range measurements from multiple GNSS constellations to determine navigation integrity. Providing robustness against multiple simultaneous satellite faults, ARAIM needs a much higher number of available measurements than a classical RAIM [1]. Studies [2, 3, 18] expect that the performance of ARAIM based integrity will be sufficient to allow for LPV-200 based operations only if two complete constellations are present. Performance simulations using GPS and Galileo constellations use an arbitrarily selected definition of the relative positioning of orbital planes of these two constellations. In reality however, the orbital plane phasing between Galileo and GPS is a determined parameter varying very slowly due to orbit perturbation. Because the RAAN (Right Ascension of the Ascending Node) parameter of all orbits drifts slowly and this drift rate depends on the orbital altitude of the spaces vehicle, Galileo and GPS have different RAAN drift rates. As a result, identical RAAN phasing between the two constellations reappears at a period of 11 years, and a potential worst case would persist for significant time, i.e. several years. Identification of such a worst case constellation phasing is thus important to avoid too optimistic performance estimates in simulations. Most previous performance studies assume that the worst case constellation phasing exists when three of the GPS planes have identical RAAN parameters to the Galileo planes as this setup fosters weak geometries where satellites from GPS and Galileo appear to be close together. This worst case assumption has been confirmed in DOP-based studies such as [4] for navigation accuracy, but not yet for ARAIM performance. Because ARAIM based navigation is much more susceptible to small and weak geometries it is necessary to review the validity of the worst case assumption with respect to ARAIM. Moreover, past work on inter-constellation phasing effects has only compared the "full alignment" scenario with the "most separated" scenario where Galileo planes are distributed exactly in the middle between the GPS planes. This paper analyses ARAIM performance for a more detailed range of RAAN phasing scenarios, and determines the worst case for ARAIM based navigation integrity. Furthermore we demonstrate the projected performance for the Galileo mission under the assumption that the recently launched Galileo SVs already define the inter-constellation phasing. By extrapolation of available orbit data to a full Galileo constellation the ARAIM performance at Galileo FOC and during the first years of operation is predicted. The results obtained from the simulations demonstrate that the constellation phasing does impact the ARAIM performance, but the magnitude of this change is small. The minor characteristic of this effect is also confirmed for a combined constellations based on current GPS configuration and Galileo at FOC. The individual impact on the ARAIM VPLs for specific users however is rather large and can be observed in both directions, i.e. the change from a "best case" to a "worst case" constellation phasing has a positive impact for the performance of some users, and a negative impact for other users