Hierarchical Deep Learning for RF Fingerprinting of Drones

Keywords

Drone, Cybersecurity, RF Signal Processing, Machine Learning

Suitable Departments

Aerospace Engineering, Computer Science, Communications Engineering, Cybersecurity, Electronics Engineering

Summary

Civilian use of drones for entertainment, delivery of emergency products, inspection and monitoring of infrastructure amongst others is increasing. Risks from drone misuse is consequently increasing. While drones must carry a MAC ID and a UIN and broadcast it regularly over a radio frequency (RF) channel, those which are being used to deliver harmful payloads, cause public disturbance or collect unauthorized footage are likely to transmit spoofed IDs or suppress their transmissions. A drone detection and classification system can expand the range of soft measures regulators can take to reduce the threat from unauthorized use of drones and facilitate the growth of businesses built around drone usage.

This PhD thesis comprises two parts: a) development and implementation of parallel, computationally efficient, hierarchical algorithms for detecting and classifying drones through intercepted communications between the drone and its controlling base station; b) empirical and theoretical studies leading to characterization of the Nash equilibrium points on latency and accuracy curves.

Programming Skills Needed

Python/Matlab C++

Hardware Skills Needed

none

Other Skills

Parallel Computing, Algorithm Design, Machine Learning, Communications Theory, Drone Communications Protocols

Partners

India : Industry
Outside : They will be arranged once you have a substantial conference publication arising out of your PhD work.

Funding

Amrita Vishwa Vidyapeetham provides stipends and teaching assistantships to selected candidates. Once you join, you could leverage existing proposals within the team to apply for additional corporate or government funding.

Time Period

3 ½ to 4 ½ years based on full-time committed

Sample References

• Anas Alsoliman, Giulio Rigoni, Marco Levorato, et. al., “COTS Drone Detection using Video Streaming Characteristics”, Intl. Conf. on Distributed Computing and Networking, Jan 2021.
• Boban Sazdic -Jotic, Ivan Pokrajac, Jovan Bajceti et. al., “Single and multiple drones detection and identification using RF based deep learning algorithm”, Expert Systems With Applications, Sep 2021.
• Dae-Il Noh, Seon-Geun Jeong, Huu-Trung Hoang et. al., “Signal preprocessing technique with noise-tolerant for RF-based UAV signal classification”, IEEE Access, Dec 2022.