A Hierarchical Framework for Drone Identification based on Radio Frequency Machine Learning

In recent years, drones have been widely used in many fields, but it also brings security threats. The detection and identification of drones, especially non-cooperative drones, is becoming a hot research topic. In this paper, we propose a hierarchical drone identification framework based on radio frequency machine learning. Our framework consists of two hierarchical stages: drone type classification and individual drone identification of the same type. In the stage of type classification, a specially designed convolutional neural network is trained to classify the signals from different types of drones. In the stage of individual identification, signals from the same type of drones are identified by estimating the pulse structure, reducing dimension through calculating fractal dimensions, and training the support vector machine classifier. In order to validate our approach, we collect and construct a dataset containing the real signals of seven DJI drones. The proposed framework achieves a classification accuracy of over 99% in both stages and performs well even at low signal-noise ratio levels.