Drone detection, classification and neutralization through artificial intelligence based RF sensing and protocol aware jamming
Mini remotely piloted aircraft systems (RPAS) are imposing threats to national security. In recent years, the threat has become increasingly vivid due to the wide availability of drones. Several illicit incidents at several security-sensitive places like airports, national campaigns and international sports events have been recorded. The small radar cross section of mini drones makes them difficult to be detected by the existing radars. Similarly, the resemblance of drone to birds makes it difficult to be detected by visual detection. On the contrary, passive RF` detection is able to detect drones from a longer range and is less dependent on weather. The aim of this doctoral research is to study and develop a RF based detector-kill chain to detect and neutralize the drone. The study will be divided into 2 parts: (i) Detection and Classification of drone signal, and (ii) Drone neutralization by protocol aware jamming. The first part of the study will deal with the detection and classification of the signal used for communication between drone and controller. A drone performs two way communication with its controller using the command and control signal, video signal and telemetry signal. Generally, the signals are frequency hopping spread spectrum (FHSS) or direct sequence spread spectrum (DSSS), operating at 2.4 GHz or 5.8 GHz ISM band. It is possible to detect the presence of drone and the operator only by sensing one of these signals using a blind energy detection technique. However, in presence of interference (e.g. WiFi or access points), it becomes very difficult to identify a drone’s signal using any existing spectrum sensing method. The first goal of this research will be to develop a sensing and classification method based on artificial intelligence to identify a drone’s signal in presence of interference. Deep learning models can be used to classify the type of modulation and further distinguish the command and control, telemetry and video signal. The sensing and classification algorithm should also be able to classify the type of drone from the time and frequency signature of the signal. Along with detection and identification, it is important to neutralize the drone for security purposes. The second part of this research will focus on drone neutralization through RF jamming. The existing RF jammer jams the complete frequency band used by the drone with a high power noise signal. It neutralizes the drone, however, creates interference to all RF communication systems in the surroundings, which makes it impossible to use in public places. The purpose of the second part of this research is to develop a smart RF jammer to neutralize the drone with a minimal interference to any system or service in the surrounding. The idea is to perform protocol aware jamming using the knowledge learnt from the classification stage with the lowest possible transmit power. Directional jamming will be performed using a directive antenna. The jamming will incorporate drone localization and tracking.