






The project aims to develop new theories and methods for enhancing the security of control systems in the presence of strategic, realistically constrained adversaries and probabilistic uncertainty. Uncertainty can play an interesting dual role in security. The adversary may seek to obfuscate their actions within the defender’s uncertainty. The defender can exploit the adversary’s lack of full knowledge and carefully inject additional uncertainty to obfuscate the system’s vulnerabilities, as in moving-target defense. Furthermore, implementation details and cross-layer interactions with communication and hardware can further affect the actions available to the adversary. We seek to develop analysis and design algorithms that incorporate cross-layer information and account for trade-offs between nominal performance, resilience to undetectable attacks, and levels of uncertainty at the operator and adversary. The exact details of the project will be decided in a dialogue between the student and supervisors. As part of an MSCA Doctoral Network, the project includes extended research visits abroad of up to 6 months in total. Planned supervisors are Assoc. Prof. André Teixeira and co-supervisor Prof. Christian Rohner.