Item

Abstract

GNSS applications for precise positioning and navigation, especially in urban environments, have been benefited by the continuous advancements of the GNSS satellite systems (GPS, GLONASS, Galileo and BeiDou). The increasing number of satellites improves positioning accuracy; however, the simultaneous use of multiple GNSS satellite systems leads to longer processing time, higher power consumption, and larger data size. Recent methods have been developed to optimise the satellite constellation by choosing only an adequate number of satellites and achieve high positioning precision. In this study, we evaluated the performance of five different artificial intelligence techniques in optimising the GNSS satellite constellation; the Artificial Bee Colony (ABC), Ant Colony Optimization (ACO), Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Simulated Annealing (SA). From the evaluation it was proved that ABC algorithm was the fastest and the most accurate in determining the optimal satellite constellation, by selecting in all cases of GNSS satellite numbers the constellation with the lowest Weighted-GDOP value. Furthermore, we made one step further by evaluating the dependency of the PPP multi-GNSS precision for the various cases of satellite constellations optimised by ABC algorithm. As it is expected the precision of PPP solution increases with the number of satellites, but the error of the PPP solution follows an exponential decay and stabilises to a value of about 20-30 mm when more than 20 GNSS satellites are used with the ABC optimisation algorithm.