GNSS-Based Solutions Testing in an ERTMS Context: A Framework for Statistical Performance Analysis

Quentin Mayolle, Juliette Marais, Martin Fasquelle, Vincent Tardif, and Emilie Chéneau-Grehalle

Peer Reviewed

Abstract:	This article proposes a new methodology to label GNSS data from multiple railway environments, based on real measurements and fusion with external sources of information (satellite and infrared). A precise attribution of each GNSS observation to one specific environment is made possible, allowing the study of short time intervals (several seconds), and the analysis of local errors, computed with the Code-Minus-Carrier method. In addition, a complete probabilistic Bayesian model is employed to characterize the errors of each environment, based on stochastic processes to consider temporal correlations between errors. This new model is then analyzed and evaluated from the acquired data, and has the ability to generate new samples for simulation purposes. Finally, a Hidden Markov Model exploits the fitted model to perform a simple detection of the environment based on the local errors calculated. Index Terms—GNSS, railway applications, multipath, Bayesian inference, statistical modeling.
Published in:	2025 IEEE/ION Position, Location and Navigation Symposium (PLANS) April 28 - 1, 2025 Salt Lake Marriott Downtown at City Creek Salt Lake City, UT
Pages:	204 - 213
Cite this article:	Mayolle, Quentin, Marais, Juliette, Fasquelle, Martin, Tardif, Vincent, Chéneau-Grehalle, Emilie, "GNSS-Based Solutions Testing in an ERTMS Context: A Framework for Statistical Performance Analysis," 2025 IEEE/ION Position, Location and Navigation Symposium (PLANS), Salt Lake City, UT, April 2025, pp. 204-213.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In