Automatic Detection of GNSS Ionospheric Scintillation Based on Extreme Gradient Boosting Technique

Abstract

Ionospheric scintillations caused by the ionospheric plasma density irregularities adversely affect the positional accuracy of the global navigation satellite system (GNSS) receiver. Machine learning methods are robust and efficient for detecting and classifying the ionospheric scintillation effects in GNSS signals. In this letter, we propose an extreme gradient boosting (XGBoost) based machine learning method to detect and classify ionospheric amplitude scintillation, applied on a large global positioning system (GPS) dataset collected from an equatorial ionization anomaly (EIA) region, Sao Jose, Brazil (geographic: 23.2°S 45.9°W; dip latitude: 20.9°S). The performance of the proposed method is compared with a classifier based on neural network (NN), support vector machine (SVM), decision tree (DT), and logistic regression (LR) methods. The confusion matrix results show that the XGBoost method performed well, with a prediction accuracy of 99.88% than other machine learning methods. XGBoost algorithm handles the data irregularities efficiently by setting a direction of descent through adaptive learning and can subsample between the columns to reduce the relevance of each weak learner. The performance results in terms of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$F1$ </tex-math></inline-formula> score, precision, recall, and area under precision-recall curve (AUC-PR) indicates that the XGBoost algorithm can characterize the ionospheric threats in GNSS signals to improve the position accuracy.