Soodeh Dadras1, Sara Dadras2, Chris Winstead1
11:00 - 11:20 | Wed 10 Jul | Room 406 | WeA15.4
In this paper, we propose a mitigation scheme to prevent an attacker from causing collisions in a vehicular platoon under a gain modification attack. A control algorithm, based on fractional-order calculus and using only local sensor information, is shown to significantly alleviate the impact of the attacker. The control is incorporated into the system when the attacker(s) is(are) detected. We prove that once the stream has been destabilized and its states continually deviate from the desired trajectories, the attacker can be interrupted by another member of the platoon. Simulations demonstrate that by applying our proposed control method, collisions are eliminated despite the attacker maliciously altering its gain such that system becomes unstable or shows oscillatory behavior.