Machine Learning Empowered Radio Frequency Signal Classification for UAS Detection
Abstract
Rapid developments in the unmanned aerial systems (UAS) have made its usage in a variety of offensive as well as defensive applications especially in military, high priority and sensitive government sites. The ability to accurately classify over-the-air radio signals will provide insights into spectrum utilization, device fingerprinting and protocol identification. These insights can aid estimating the UAS transmitters capabilities without their knowledge. In this paper, we present a Radio Frequency Signal Classification (RF-Class) toolbox that can monitor, detect, and classify wireless signals emitted by UAS. The advantage of the RF-Class toolbox is extracting information about transmitters and providing receivers information about certain transmitted signals. The classification of RF signals will be done based on the modulation scheme recognition, exploitation of cyclostationary features and leveraging RF band allocation information. The modulation recognition capability can also be used for cyber offensive strategies. Once the modulation scheme is recognized, we can demodulate, decode and extract packets. Once the packets are extracted, we can accurately detect the protocol. The final step involves crafting a malicious packet and injecting the packet in the adversarial communication environment with intent to launch offensive operations. To demo the feasibility and accuracy of our approach, we have evaluated the performance on a real environment with an UAS (Drone - DJI Phantom 4). Our initial experimental result showed that we were able to detect presence of drone signal successfully in presence of varying SNR regimes.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 30, 2021
- Accession Number
- AD1152146
Entities
People
- Charles Kamhoua
- Kimberly Gold
- Michael Nilsen
- Sachin Shetty
Organizations
- Naval Surface Warfare Center
- United States Army Research Laboratory