Abstract:This paper develops a glocal (global-local) attack detection framework to detect stealthy cyber-physical attacks, namely covert attack and zero-dynamics attack, against a class of multi-agent control systems seeking average consensus. The detection structure consists of a global (central) observer and local observers for the multi-agent system partitioned into clusters. The proposed structure addresses the scalability of the approach and the privacy preservation of the multi-agent system's state information. The former is addressed by using decentralized local observers, and the latter is achieved by imposing unobservability conditions at the global level. Also, the communication graph model is subject to topology switching, triggered by local observers, allowing for the detection of stealthy attacks by the global observer. Theoretical conditions are derived for detectability of the stealthy attacks using the proposed detection framework. Finally, a numerical simulation is provided to validate the theoretical findings.

What problem does this paper attempt to address?

There are two main problems that this paper attempts to solve: 1. **Achieving Average Consistency under Privacy Protection**: In a multi - agent control system, how to ensure that the system does not leak the initial state of the system or the final consensus value (i.e., \( x^*=\frac{1}{N}\sum_{i = 1}^{N}x_i(t_0) \), \( v^* = 0 \)) during the process of achieving average consistency, and at the same time ensure that the communication topology of the system cannot be reconstructed. 2. **Scalable Stealthy Attack Detection**: For multi - agent systems suffering from stealthy attacks (such as covert attacks and zero - dynamic attacks), how to design a decentralized and scalable attack - detection framework to ensure that these attacks can be effectively detected, and this framework meets the above - mentioned privacy - protection requirements. ### Specific Problem Description #### 1. Achieving Average Consistency under Privacy Protection - **Problem Background**: Nodes in a multi - agent control system need to achieve average consistency through local information exchange. However, in this process, the initial state of the system, the final consensus value, and the communication topology may be exploited by malicious attackers. - **Objective**: Design a method so that the initial state and final consensus value of the system will not be leaked or reconstructed, and at the same time, the communication topology of the system cannot be reconstructed. #### 2. Scalable Stealthy Attack Detection - **Problem Background**: Stealthy attacks (such as covert attacks and zero - dynamic attacks) are difficult to detect by traditional observer methods because they are invisible in the system output. In addition, the distributed nature of large - scale network systems makes centralized detection methods difficult to implement. - **Objective**: Develop a decentralized and scalable attack - detection framework that can detect stealthy attacks without leaking privacy and has good scalability. ### Main Contributions of the Paper 1. **Privacy Protection**: Propose to impose unobservability constraints on the network topology to protect the privacy of the initial conditions and final consensus values of nodes. 2. **Scalable Attack - Detection Structure**: Propose a global - local (glocal) attack - detection structure, divide the multi - agent system into multiple subsystems (clusters), and each subsystem has its own global and local observers to meet the needs of privacy protection and attack detection. 3. **Topology Switching Strategy**: Derive the theoretical conditions for topology switching, so that the local detector can trigger the switching of the communication topology to make stealthy attacks detectable in the global observer. ### Conclusion This paper solves the problems of stealthy attack detection and privacy protection in multi - agent control systems by introducing a privacy - protection and scalable attack - detection framework. The proposed global - local attack - detection structure and topology - switching strategy provide new ideas and methods for future related research.

Privacy-Preserving Stealthy Attack Detection in Multi-Agent Control Systems

Design of Strictly Stealthy Attacks for Nonlinear Multiagent Systems under Distributed Attack Detection Mechanism

Stealthy Actuator Signal Attacks in Stochastic Control Systems: Performance and Limitations.

Novel Stealthy Attack and Defense Strategies for Networked Control Systems

Undetectable Cyber Attacks on Communication Links in Multi-Agent Cyber-Physical Systems

Cyber-Attack Detection in Discrete Nonlinear Multi-Agent Systems Using Neural Networks

Cooperative control for cyber-physical multi-agent networked control systems with unknown false data-injection and replay cyber-attacks

Detection of Stealthy Adversaries for Networked Unmanned Aerial Vehicles*

Application of the unified control and detection framework to detecting stealthy integrity cyber-attacks on feedback control systems

Privacy-Preserving Adaptive Resilient Consensus for Multi-Agent Systems under Cyber Attacks

Risk Assessment for Nonlinear Cyber-Physical Systems under Stealth Attacks

Stealthy Attacks for Partially-Observed Discrete Event Systems

Cyberattack Detection for a Class of Nonlinear Multiagent Systems Using Set-Membership Fuzzy Filtering

Optimal Strictly Stealthy Attack Design on Cyber-Physical Systems: A Data-Driven Approach.

Interval observer-based consensus tracking for multi-agent systems under Markovian switching topologies with stealthy attack

An Online Kullback-Leibler Divergence-Based Stealthy Attack Against Cyber-Physical Systems

Event-Triggered Secure Control of Nonlinear Multi-agent Systems under Sensor Attacks

Cyberattack Detection for Nonlinear Leader-Following Multi-Agent Systems Using Set-Membership Fuzzy Filtering

Attack Detection for Cyber-Physical Systems: A Zonotopic Approach

Adversary Detection and Resilient Control for Multi-Agent Systems

Performance Analysis for Cyber–physical Systems under Two Types of Stealthy Deception Attacks