Resilient Autonomous Air and Space Systems
Abstract
To maintain dominant tactical autonomy, uncrewed air and space robotic vehicles must be able to detect and respond to potential attacks on the sensor, telemetry, and actuation signals in their systems. Classical cybersecurity mechanisms cannot prevent analog and interference attacks. Therefore, we must develop new attack mitigationstrategiesthatleveragedata-drivensensorinformationandsimulationmodelstodetect, respond, and design resilient systems. To fill this gap, this project proposes the first systematic study of how to design data-driven defenses to improve the resiliency of airborne and spaceborne vehicles. We will leverage new information processing techniques to analyze spatiotemporal signals in real-time. We will leverage high-fidelity models of the vehicle and the physical- world to complement these signals and design appropriate attack-response strategies. In Task1, we will focus on improving our situational awareness after an attack. In Task2, we will develop data-driven mitigation algorithms to survive the attack and either complete the mission or return to a safe state. Finally, in Task3, we will work on simulation-based exploration of potential vulnerabilities to identify problems early on and design more resilient airborne and spaceborne remote or uncrewed robotic vehicles. Our work to detect and mitigate these attacks will fill the state-of-the-art gap and help maintain full spectrum superiority in uncrewed and robotic operations.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 05, 2025
- Source ID
- FA95502410015
Entities
People
- Alvaro Cardenas
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of California, Santa Cruz