Situational Awareness of Chirp Jamming Threats to GNSS Based on Supervised Machine Learning

The presence of malicious jamming attacks is disruptive for global navigation satellite system (GNSS) applications requiring high robustness. The situational awareness aims to timely detect the jamming attacks and characterize the information of the jamming signals for further response. Recently, machine learning (ML) techniques have been extensively explored in GNSS applications as a means to detect outliers at signal level. This article addresses the detection and classification of chirp jamming signals through
$k$
-nearest neighbor techniques applied at the precorrelation stage. The algorithm is able to detect the presence and classify the power strength and sweep rate of the chirp signals, thus enabling the possibility to properly tune the mitigation of the interference. Compared to the traditional techniques for detection and classification of chirp signals which usually require postprocessing and human-driven analysis, the proposed method based on ML can automatically detect and characterize the chirp signals, showing the potential to be applied in the scenarios where timely awareness and response to the jamming attacks are in demand.