GNSS Signal Spoofing Detection based on Satellite Oscillator Stability

Global navigation satellite systems (GNSS) provide all-weather timing, positioning and navigation services, playing a crucial role in both military and civilian applications. Despite their widespread use, GNSS is vulnerable to severe signal spoofing attacks. In this paper, we propose a novel GNSS signal spoofing detection method by exploiting the radio frequency fingerprint (RFF) of GNSS satellites. We introduce a new RFF feature named satellite oscillator stability, which is closely related to the hardware impairments of GNSS satellites and extracted from the GNSS receiver carrier tracking loop. Utilizing this new feature, we propose a GNSS signal spoofing detection method, which takes the extracted satellite oscillator features as input and employs a Residual Network as a classifier to distinguish between legitimate and spoofed signals. For verification, we use the most widely used GPS and Beidou signals, and conducted experiments in two scenarios: GPS generative spoofing and Beidou forwarding spoofing. The experimental results demonstrated that our proposed method achieve a detection accuracy of over 98% in the GPS generative spoofing scenario, and over 93% in the Beidou forwarding spoofing scenario.