MULTIVARIATE SIMULATION AND VALIDATION OF GNSS OBSERVATIONS

Abstract

This research explores the simulation and validation of GNSS observations using computer tools. A 24-hour set of observables from day 354 of 2019, belonging to the HAVA continuous tracking station, was simulated. The simulation was run using Bernese v5.2 scientific software, specifically its GPSSIM program, generating a synthetic observation file in RINEX format that compiles different code and phase observables available in the GPS and GLONASS constellations at the time of the simulation. The results of this simulated observation were subsequently processed using the AUSPOS online post-processing service, operated by Geoscience Australia, to obtain an accurate coordinate solution.
The comparison between the coordinates obtained from the simulation and the reference values ??from an actual observation at the HAVA station revealed a three-dimensional discrepancy of 2.4 centimeters. To assess the statistical significance of this difference, Hotelling's T² test, a multivariate hypothesis testing tool, was applied. The statistical analysis did not find sufficient evidence to reject the null hypothesis of equality between the coordinate vectors, with a p-value of 0.47, indicating that the observed differences are not statistically significant. It is concluded that the simulation reproduces the actual observations with high fidelity at the centimeter level.