Optimal DoS attack and proactive deception strategies on remote state estimation: Utilizing subsystem importance

Abstract

This paper tackles security challenges in remote state estimation for cyber–physical systems (CPSs) with a focus on an importance-based denial-of-service (DoS) attack over an infinite time horizon. We employ the concept of “importance” to quantify the criticality of subsystem information for decision efficiency, establishing a correlation between importance and transmission power allocated to each subsystem’s channel. We establish the existence of an optimal attack policy and analyze its structural properties. Additionally, we investigate proactive deception strategies that manipulate transmission power to mislead attackers about the importance of subsystems. Within a Stackelberg game framework, we propose and analyze a proactive deception strategy tailored against importance-based DoS attacks, along with an algorithm to obtain the optimal strategy. Finally, we provide numerical examples to illustrate the theoretical results and demonstrate the practical implications of our proposed strategies for enhancing security in remote state estimation for CPSs.