This is the first of two papers authored by LEFT Lab PhD student Soodeh Dadras which appeared in the proceedings of the 2018 American Controls Conference. #^Identification of the Attacker in Cyber-Physical Systems with an Application to Vehicular Platooning in Adversarial Environment - IEEE Conference Publication
Cyber-Physical Systems (CPS) are systems with tight coupling between integration of physical, computational and networking components. Control systems play an important role to help these systems to adhere to their desired performance. Having a reliable and secure control system which can cope with high risk situations and various attacks is one of the bottlenecks for real world cyber-physical systems. It has been proven that using control modification attack, where the adversary modifies the sensor information or the control law, can disrupt the desired performance of the system. In this paper, a novel scheme is presented to detect and identify the attacker in control systems. The detection algorithm is the combination of the system identification method and machine learning technique which effectively recognizes the malicious actors. The proposed algorithm is efficient, viable, and simply adequate to address the challenges posed by complex cyber-physical systems. Finally, the efficiency of the presented method is verified with the case of platooning in an adversarial environment.