@ericyocam6Professor, California Polytechnic State University
Shipped: TriSweep: Drone Swarm EM Side-Channel Analysis
A four-drone swarm that autonomously sweeps a target and captures electromagnetic side-channel leakage from hardware, turning EM analysis into a mobile, aerial capability.
dronesswarm-roboticsside-channelhardware-securityelectromagnetic
Comments2
Four drones with SDRs could give you spatial diversity, but keeping them phase and clock coherent in flight seems like the hard part. Are you combining captures across drones, or treating each as an independent probe? I'd also like to know how motion and platform noise affect trace alignment.
Great question, Ryan. TriSweep uses coherent signal combining across the four drones rather than treating each as an independent probe. Our simulation demonstrates a +6.0 dB SNR improvement using four drones, with a 90-degree swarm configuration supporting 3D source localization. You're correct that maintaining clock and phase coherence in flight is a significant engineering challenge, particularly with motion, vibration, and onboard electrical noise. Our current proof of concept validates the combining model using ASCAD traces and simulated EM conditions. Physical synchronization and real-world trace alignment remain important next steps in validating the architecture.
2 comments
Four drones with SDRs could give you spatial diversity, but keeping them phase and clock coherent in flight seems like the hard part. Are you combining captures across drones, or treating each as an independent probe? I'd also like to know how motion and platform noise affect trace alignment.
Great question, Ryan. TriSweep uses coherent signal combining across the four drones rather than treating each as an independent probe. Our simulation demonstrates a +6.0 dB SNR improvement using four drones, with a 90-degree swarm configuration supporting 3D source localization. You're correct that maintaining clock and phase coherence in flight is a significant engineering challenge, particularly with motion, vibration, and onboard electrical noise. Our current proof of concept validates the combining model using ASCAD traces and simulated EM conditions. Physical synchronization and real-world trace alignment remain important next steps in validating the architecture.