RF Fingerprint Measurement For Detecting Multiple Amateur Drones Based on STFT and Feature Reduction
Chengtao Xu, Bowen Chen, Yongxin Liu, Fengyu He, Houbing Song
Abstract
Underlying the easy accessibility and popularity of amateur unmanned aerial vehicles (UAVs, or drones), an effective multi-UAV detection method is desired. In this paper, we proposed a novel radio frequency (RF) signal detection method for recognizing multiple UAVs’ intrusion. The single transient control and video signal is transformed by Short Time Fourier Transform (STFT) to obtain its time-frequency-energy distribution features. To reduce the dimensionality of the RF feature vector, the principal component analysis (PCA) is applied in the signal characteristic subspace transformation. A remapped UAVs RF signal feature data is used in the training of the support vector machine (SVM) and K-nearest neighbor (KNN) algorithm for classifying the presence and number of intruding UAVs. In addition, a real-time test of UAV attacks on an airport area is implemented. The test results show that the accuracy for detecting the number of intruding UAVs is effective. This method could similarly apply to protect the public from unsafe and unauthorized UAV operations near security sensitive facilities.