A Quadratic-Spectrum-Based Radio Frequency Fingerprinting Method for Unmanned Aerial Vehicles
Tianjia Li, Yuelei Xie, Xiangguo Liu · Advances in transdisciplinary engineering · 2023
With the increase of unmanned aerial vehicle (UAV) application scenarios, the incidents of UAV interference and malicious harassment are increasing, which pose a threat to public privacy and security. Aiming at the problem that UAVs are difficult to control, this paper proposes a radio frequency fingerprinting (RFF) method based on the quadratic spectrum. Firstly, the small window power detection is used to detect the transmission signal of the image to be recognized and extract the effective signal. The quadratic spectrum preprocessing is performed on the extracted signal, and finally, the improved VGG-16 deep convolutional neural network in this paper is used to extract the RFF information of the signal. The softmax classifier is used to classify and recognize the signal. The experiment is carried out in three steps, and the results prove that the combination of quadratic spectrum and improved VGG-16 deep convolutional neural network for RFF can complete a series of identification of “whether – what-which”, and the time consumed is shorter.