Primary Investigator:
Inseok Hwang

Project Members
Sounghwan Hwang, Minhyun Cho, and Inseok Hwang

Unmanned Aerial Vehicles (UAVs) are complex cyber-physical systems that depend heavily on sensor feedback for precise control. However, these sensors are susceptible to cyberattacks, which can undermine the safety and reliability of UAV operations. Among these, stealthy sensor attacks are particularly hazardous, as they can bypass detection while incrementally degrading control performance. Motivated by this vulnerability, this study proposes a risk assessment framework for UAV tracking control under stealthy attack conditions. By characterizing the impact of attack signals on closed-loop tracking dynamics, we derive an ellipsoidal over-approximation of the reachable set using Linear Matrix Inequality (LMI)-based techniques. This reachable set identifies the possible state space the UAV might occupy when under attack. Furthermore, this framework allows for a systematic evaluation of safety risks; specifically, it can determine whether a UAV might deviate into prohibited zones or collide with obstacles in urban environments. Consequently, the proposed approach serves as a proactive safety analysis tool for UAV operations in adversarial settings.