Stackelberg-game analysis of correlated attacks in cyber-physical systems

This paper studies a resilient control problem for discrete-time, linear time-invariant systems subject to state and input constraints. State measurements and control laws are transmitted over a communication network and could be corrupted by human adversaries. In particular, we consider a class of human adversaries, namely correlated jammers, who are modeled as rational decision makers and whose strategies are highly correlated to the control system operator. The coupled decision making process is modeled as a two-level receding-horizon dynamic Stackelberg (leader-follower) game. We propose a receding-horizon Stackelberg control law for the operator, and analyze the resulting performance and closed-loop stability of the system under correlated attacks. We observe that, with full information of his follower, the operator is still able to maintain regional stability of the control system.