Evaluation of (Un-)Supervised Machine-Learning-Based Detection, Classification, and Localization Methods of GNSS Interference in the Real World

Feigl T, Brieger T, Ott F, Hansen J, Contreras DF, Ruegamer A, Felber W (2023)


Publication Language: English

Publication Type: Conference contribution, Original article

Publication year: 2023

Pages Range: 1-13

Conference Proceedings Title: Proc. Intl. Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+)

Event location: Denver, CO US

Abstract

The accuracy and reliable localization of vehicles on roads is crucial for self-driving cars, toll systems, and digital tachographs. To achieve this, vehicles typically employ Global Navigation Satellite System (GNSS) receivers to validate their absolute position information. However, GNSS-based positioning can be compromised by interferences signals, necessitating the identification, classification, purpose determination, and localization of the interference to mitigate or eliminate it. Recent approaches based on artificial intelligence (AI) have shown superior performance in interference monitoring. However, their feasibility in real applications and environments is yet to be determined. To effectively implement (supervised) AI techniques, it is necessary to have (sensor) training datasets that incorporate realistic interference signals, including real-world noise and potential multipath effects that may occur between transmitter, receiver, and satellite in the operational area. Additionally, these datasets require reference labels. However, creating such datasets is often challenging due to legal reasons, as the use of GNSS sources of interference is strictly prohibited. The evaluation of AI-based methods in the literature has been limited to either synthetic data or controlled laboratory environments. Consequently, the performance of AI-based methods in practical applications remains unclear. To address this gap in knowledge, this paper describes a series of large-scale measurement campaigns conducted in the real world with special permission from legal authorities to compensate for limitation. To evaluate the applicability of AI-based techniques in detecting, classifying, and locating GNSS interference in real-world scenarios, we conduct a baseline experiment in an anechoic chamber. We also conduct experiments in the Seetal Alps in Austria, and a large-scale, long-term measurement camping at two Autobahn spots in Germany. We evaluate the latest (un)supervised AI-based techniques to report on their performance in real-world settings.

Authors with CRIS profile

Involved external institutions

How to cite

APA:

Feigl, T., Brieger, T., Ott, F., Hansen, J., Contreras, D.F., Ruegamer, A., & Felber, W. (2023). Evaluation of (Un-)Supervised Machine-Learning-Based Detection, Classification, and Localization Methods of GNSS Interference in the Real World. In Proc. Intl. Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+) (pp. 1-13). Denver, CO, US.

MLA:

Feigl, Tobias, et al. "Evaluation of (Un-)Supervised Machine-Learning-Based Detection, Classification, and Localization Methods of GNSS Interference in the Real World." Proceedings of the Proc. Intl. Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+), Denver, CO 2023. 1-13.

BibTeX: Download