In this paper, we present a bagged tree model able to detect phase scintillation at high latitudes with 95% accuracy, 5% scintillation miss-detection and 5% scintillation false alarm. The input to the model is a series of 3 minutes of the Total Electron Content (TEC), 3 minutes of the change in TEC (dTEC), and the satellite elevation. These values are extracted from Ionospheric Scintillation Monitoring Records (ISMR) logged by Ionospheric Scintillation Monitoring (ISM) receivers. We compare the performance of this model to Support Vector Machine (SVM) models, k-Nearest Neighbors (k-NN) models, and also to other decision tree models. Furthermore, we assess the ability of the TEC and dTEC features to detect scintillation independently of the scintillation indexes. For this, we compare the above decision trees, kNN and SVM models to the same models but trained using scintillation indexes as additional inputs. Moreover, we show the results of testing the proposed model using a novel data set. Finally, we compare the accuracy of the machine learning model to the performance of a detector based on the phase scintillation index σ ϕ threshold.

Detecting Phase Scintillation at High Latitudes Using Ionospheric Scintillation Monitoring Records and Machine Learning Techniques / Imam, Rayan; Savas, Caner; Dovis, Fabio. - ELETTRONICO. - (2021), pp. 37-42. ((Intervento presentato al convegno 2021 IEEE International Conference on Wireless for Space and Extreme Environments (WiSEE) tenutosi a Cleveland, OH, USA nel 12-14 Oct. 2021 [10.1109/WiSEE50203.2021.9613840].

Detecting Phase Scintillation at High Latitudes Using Ionospheric Scintillation Monitoring Records and Machine Learning Techniques

Imam, Rayan;Savas, Caner;Dovis, Fabio
2021

Abstract

In this paper, we present a bagged tree model able to detect phase scintillation at high latitudes with 95% accuracy, 5% scintillation miss-detection and 5% scintillation false alarm. The input to the model is a series of 3 minutes of the Total Electron Content (TEC), 3 minutes of the change in TEC (dTEC), and the satellite elevation. These values are extracted from Ionospheric Scintillation Monitoring Records (ISMR) logged by Ionospheric Scintillation Monitoring (ISM) receivers. We compare the performance of this model to Support Vector Machine (SVM) models, k-Nearest Neighbors (k-NN) models, and also to other decision tree models. Furthermore, we assess the ability of the TEC and dTEC features to detect scintillation independently of the scintillation indexes. For this, we compare the above decision trees, kNN and SVM models to the same models but trained using scintillation indexes as additional inputs. Moreover, we show the results of testing the proposed model using a novel data set. Finally, we compare the accuracy of the machine learning model to the performance of a detector based on the phase scintillation index σ ϕ threshold.
978-1-6654-0371-9
File in questo prodotto:
File Dimensione Formato  
WiSEE21_conferenceFullPaper_pdfexpress.pdf

accesso aperto

Descrizione: Final version of the Submitted manuscript.
Tipologia: 2. Post-print / Author's Accepted Manuscript
Licenza: PUBBLICO - Tutti i diritti riservati
Dimensione 797.59 kB
Formato Adobe PDF
797.59 kB Adobe PDF Visualizza/Apri
Detecting_Phase_Scintillation_at_High_Latitudes_Using_Ionospheric_Scintillation_Monitoring_Records_and_Machine_Learning_Techniques.pdf

non disponibili

Descrizione: Online version
Tipologia: 2a Post-print versione editoriale / Version of Record
Licenza: Non Pubblico - Accesso privato/ristretto
Dimensione 1.31 MB
Formato Adobe PDF
1.31 MB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2951313