GNSS Satellite Selection Algorithm Revisited: A Weighted Way with Integrity Consideration

Abstract

With the promising future of Global Navigation Satellite System (GNSS), the quantity of available satellites could dramatically increase to more than 35. Tracking all available satellites simultaneously leads to unacceptable computation burden for standalone receivers. Most satellite selection methods were therefore proposed to release the computation load with the minimum satellites set to maintain endurable positioning accuracy under the least Dilution of Precision (DOP) loss criterion. In terms of safety-of-life (SoL) navigation services, the integrity monitoring is an indispensable function. The traditional satellite selection methods should therefore be enhanced to take account of positioning accuracy and the potential integrity loss from some fault satellite, which may originates from satellite clock anomaly drift or the deformation of signals in space, e.g. ionosphere scintillation. Giving the fact that the loss of some satellite may have little difference on the DOP, but could impose a negative influence on the integrity performance, satellite selection should therefore give an insight into the selection criterion and the determination of the threshold values. A novel satellite selection algorithm with using weighting method is presented to select the optimal satellites set with satisfying the requirement in both accuracy and integrity. The selection threshold values for DOP and integrity is deduced from the required navigation performance (RNP) defined by International Civil Aviation Organization (ICAO). The performance of the proposed method was validated with using the simulated GNSS constellations. The accuracy and integrity also were compared with the typical existing satellite selection methods, i.e., maximum volume algorithm and quasi-optimal satellite selection algorithm, etc. The simulation results demonstrate that weighted satellites selection method could provide optimal satellites set with not only the better accuracy and less integrity loss, but also with a reasonable computation burden, and thus could be an edge option for the GNSS satellite selection methods.