Radio frequency fingerprint-based unmanned aerial vehicle detection method
Li Wei · 2023
With unmanned aerial vehicle (UAV) technology advancing at a rapid pace, the accurate monitoring and detection of UAVs have become increasingly crucial. Traditional methods relying on radar, vision, and optical tracking may fail to detect small-sized UAVs. This research proposes an architecture for unmanned aircraft signal radio frequency fingerprint (RFF) recognition based on Empirical Mode Decomposition (EMD). The architecture retains certain Intrinsic Mode Functions (IMFs) to describe the RFF of the signal, utilizing deep learning techniques to extract fingerprint features and perform signal recognition. Validation of the proposed architecture was conducted using ADS-B signals. Experimental results demonstrate that, compared to other traditional machine learning recognition methods, the proposed architecture achieves higher recognition accuracy in both high and low signal-to-noise ratio environments. This study provides the strong technical backbone for the potential development of ever more robust and advanced UAV detection systems, potentially with far-reaching implications for ensuring UAV aeronautical safety and detection of the electromagnetic environment.