Session A1: Navigation Security and Authentication

Performance Evaluation of Specific Force-Aided Spoofing Detection Method in Complicated Scenarios
Yu Wang, Chao Sun, Bin Yuan, Lu Bai, Kaiyan Jin, Yingzhe He, Shuai Zhang, Jiayue Lei, Department Electronic and Information Engineering, Beihang University
Location: Holiday 1 (Second Floor)

Global navigation satellite systems (GNSS) are increasingly vulnerable to spoofing attacks, which can mislead receivers into reporting inaccurate navigation solutions and pose severe threats to safety-critical applications. While inertial navigation systems (INS) enhance the robustness of GNSS, conventional innovation-based spoofing detectors suffer from high false alarm rates when the user equipment (UE) undergoes maneuvers. In our previous work, we proposed a specific force-aided spoofing detection method with an F-distribution-based test statistic, which effectively reduces maneuver-induced false alarms. However, extensive evaluations reveal degraded spoofing detection performance, particularly when spoofing and maneuvers occur simultaneously. To address this issue, we propose a novel, specific force-aided spoofing detection framework that combines traditional statistical methods with deep learning techniques. First, we develop a data-fitting-based predictor to estimate the post-maneuver EKF divergence duration during which the conventional innovation-based spoofing test statistic remains above the threshold after a true maneuver, further mitigating false alarms. Second, we design a long short-term memory (LSTM) detection module to detect spoofing during the pseudo-maneuver period, which includes both the actual maneuver duration and the post-maneuver extended Kalman filter (EKF) divergence duration. Finally, a state-adaptive detection strategy uses accelerometer and gyroscope measurements to determine whether the UE is maneuvering, and switches between the conventional innovation-based and LSTM-based detection modules. Simulation results show that the proposed method significantly reduces the false alarm rate caused by UE maneuvers and improves spoofing detection performance in scenarios where spoofing and maneuvers occur simultaneously.