Detectability of Intermittent Zero-Dynamics Attack in Networked Control Systems

Abstract

This paper analyzes stealthy attacks, particularly the zero-dynamics attack (ZDA) in networked control systems. ZDA hides the attack signal in the null-space of the state-space representation of the control system and hence it cannot be detected via conventional detection methods. A natural defense strategy builds on changing the null-space via switching through a set of topologies. In this paper, we propose a realistic ZDA variation where the attacker is aware of this topology-switching strategy, and hence employs the policy to avoid detection: "pause (update and resume) attack" before (after) topology switching to evade detection. We first systematically study the proposed ZDA variation, and then develop defense strategies under the realistic assumptions. Particularly, we characterize conditions for detectability of the proposed ZDA variation, in terms of the network topologies to be maintained, the set of agents to be monitored, and the measurements of the monitored agents that should be extracted. We provide numerical results that demonstrate our theoretical findings.

Novel Stealthy Attack and Defense Strategies for Networked Control Systems

“…where x i (t) ∈ R is the position, v i (t) ∈ R is the velocity, and u i (t) ∈ R is the local control input 1 . The broad applications of its coordination control is the main motivation of this paper considering the model (1). For coordination control, we consider more representative average consensus.…”

“…2) cooperatively work with a stealthy topology attack, such that the original ZDA policy continues to be feasible under the corrupted topology (cooperative ZDA). In [1], we present our preliminary results on the conditions for detectability of an intermittent ZDA. In this paper, building on [1], we develop integrated defense strategies for both intermittent and cooperative ZDA variants, in the presence of privacy considerations.…”

“…In [1], we present our preliminary results on the conditions for detectability of an intermittent ZDA. In this paper, building on [1], we develop integrated defense strategies for both intermittent and cooperative ZDA variants, in the presence of privacy considerations. More specifically, we develop defense strategies to address the following questions: what network topology should be maintained, which agents should be monitored and what measurements the monitored agents should output, such that the intermittent and cooperative ZDA variants are detectable, and at the same time, the privacy of nonmonitored agents' real-time states are preserved?…”

“…Several real-world networked systems that can be represented by(1) including the second-order consensus[9], flocking[38], swarming[39], velocity synchronization and regulation of relative distances[40], and their applications in the decentralized formation control of mobile robots[41] and spacecraft[42], and distributed continuous-time optimization[43], etc.…”