Characterization of GNSS Receiver Position Errors for User Integrity Monitoring in Urban Environments

International audience; The characterization of GNSS position errors in urban environments is an important issue for integrity monitoring and classification of receivers’ performance. However, these errors are not observable directly by the receiver, therefore RAIM methods use statistics based on the pseudorange residuals (i.e. observable errors). In this work, we focus on the modelling and analysis of navigation errors in the position-domain rather than individual range-domain errors that are difficult to model in urban environments due to multipath and non-line-of-sight (NLOS) signals. Using a trajectory of reference we compute the horizontal position errors (HPE) and its non-parametric distribution function given by the empirical Cumulative Distribution Function (CDF). According to the results, we observe that these errors have a heavy-tailed distribution, and then we propose to fit the empirical CDF with the CDF of the generalized Pareto distribution (GPD). We use an inflated version of the fitted Pareto model to overbound the CDF of the HPE for the calculation of Horizontal Protection Level (HPL), i.e., bounding the radial position errors.