Physical Intrusion Games—Optimizing Surveillance by Simulation and Game Theory

The protection of cyber-physical networks is a topic of increasing importance. The evolution of IT (cyber) systems that control and supervise the underlying physical system has grown over decades, whereas security has not become a concern until quite recently. Advanced persistent threats (APTs) have proven to be a difficult but significant challenge for practitioners. This paper adopts a game-theoretic modeling of APTs and applies it to the (sub) problem of physical intrusion in an infrastructure. The gap between defining a good theoretical model and practically instantiating it is considered in particular. The model description serves to illustrate what is needed to put it into practice. The main contribution of this paper is the demonstration of how simulation, physical understanding of an infrastructure, and theoretical methods can be combined toward a practical solution to the physical intrusion avoidance problem. Numerical results are given to show how the physical intrusion game is being set up, and how the results obtained from its analysis can be interpreted and used for an optimized defense.