CRII: SaTC: MUSE: MUlti-modal Security Enhancement for Cyber Resilient UAV Operation

Unmanned Aerial Vehicles (UAVs), commonly known as drones, are increasingly used in critical applications such as emergency response, infrastructure inspection, and environmental monitoring. However, these systems are highly vulnerable to cyber threats, especially global positioning system (GPS) spoofing attacks that can mislead navigation and cause serious consequences. The project’s novelties are the development of an AI-powered, GPS-independent navigation and localization framework that enables UAVs to operate securely and autonomously in GPS-denied or adversarial environments. The project's broader significance and importance are in contributing to a secure and trustworthy cyber-physical infrastructure, enhancing preparedness, and improving the resilience of essential aerial services. This project also includes outreach activities and training opportunities for undergraduate, graduate, and high school students in cybersecurity. The research focuses on designing a MUlti-modal Security Enhancement (MUSE) framework that integrates sensor data from Light Detection and Ranging (LiDAR) sensors, cameras, and other motion sensors to achieve real-time localization and spoofing detection without relying on GPS. The technical approach involves building a multi-modal localization module and a cyber-resilient path planning system that can guide UAVs in GPS-denied or spoofed environments. These modules are evaluated on a Multi-Modal AI-enabled UAV Research (MM-AIR) testbed, enabling validation under diverse operating conditions. The project contributes novel methods in sensor fusion, trajectory reconstruction, and secure autonomous navigation, with expected impact across the fields of cyber-physical systems, AI, and cybersecurity. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.