A Machine Learning Approach to GNSS Scintillation Detection: Automatic Soft Inspection of the Events

Alfredo Favenza, Alessandro Farasin, Nicola Linty, and Fabio Dovis

Abstract:	Classical approaches for the automatic detection of ionospheric scintillation events in Global Navigation Satellite System (GNSS) receivers are based on the observation of indices (e.g. S4) that are obtained by processing parameters assessed at the signal processing stages of the receiver. Such values are the result of algorithms that imply specific processing choices (such as detrending, averaging and threshold operations) which influence the final performance of the detection. To reach good levels of accuracy and generalization for the identification and classification of the physical phenomenon, these approaches may require an additional human effort to refine the detection results by means of a manual inspection of the events, which is expensive and time consuming. This paper proposes a new methodology for the detection of ionospheric scintillation events based on Machine Learning techniques applied to GNSS data. This method, based on Decision Trees algorithms, aims at overcoming the limitation of the classical approaches by identifying scintillation events “as if” done by a human operator through visual inspection. This approach is automatic, unbound from traditional scintillation indices and features improved detection, false alarm, and missed detection rates when compared to standard methods.
Published in:	Proceedings of the 30th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2017) September 25 - 29, 2017 Oregon Convention Center Portland, Oregon
Pages:	4103 - 4111
Cite this article:	Favenza, Alfredo, Farasin, Alessandro, Linty, Nicola, Dovis, Fabio, "A Machine Learning Approach to GNSS Scintillation Detection: Automatic Soft Inspection of the Events," Proceedings of the 30th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2017), Portland, Oregon, September 2017, pp. 4103-4111. https://doi.org/10.33012/2017.15351
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In