A host takeover game model for competing malware

Malware, or malicious software, degrades the performance of cyber-physical systems by infecting cyber systems and compromising the information exchange between cyber and physical components. Advanced malwares have the ability to modify their code over time to escape detection, while also removing competing malwares from a targeted host. In this paper, we model the interaction of multiple adaptive, competing malwares and a system owner via a resource takeover game known as FlipIt, or a game of “stealthy takeover.” We characterize the unique Nash equilibrium of a generalization of FlipIt with an arbitrary number of players. We then prove that, by greedily updating their strategies at each time step using only local information, the malwares will converge to a unique Nash equilibrium. In addition, we derive the optimal mitigation strategy against competing malwares as the solution to a Stackelberg game and develop an efficient algorithm for computing the equilibrium. Our results are demonstrated via a numerical study, in which we analyze the behavior of the malwares prior to convergence to the equilibrium and compare the impact of heterogeneous and homogeneous malwares on the system owner´s utility.