Resilience Analysis of Closed-Loop Multiagent Systems Under Replay Attacks in the Context of Consensus Tracking

Area of Science:

• Control Systems Engineering
• Cybersecurity
• Networked Systems

Background:

• Extensive research exists on network attacks targeting control systems.
• Limited research addresses the resilience of multiagent systems (MASs) against these attacks, especially replay attacks.

Purpose of the Study:

• To propose a dynamic stability analysis method for assessing the resilience of strict-feedback nonlinear MASs under replay attacks.
• To develop a resilience metric and algorithms for analyzing tracking error bounds during aperiodic replay attacks.

Main Methods:

• Utilized a distributed adaptive consensus control framework.
• Designed a dynamic compact set model as a resilience metric.
• Employed Lyapunov stability analysis to derive tracking error bounds during attacks.

Main Results:

• Developed an iterative algorithm to compute the upper bound of tracking error jumps.
• Derived an upper bound for tracking error variation during attacks when control signals are unmodeled.
• Proved that MAS resilience is maintained if the resting time between replay attacks meets a sufficient condition.

Conclusions:

• The proposed dynamic stability analysis method effectively evaluates MAS resilience against replay attacks.
• Simulation results validate the effectiveness of the developed resilience analysis method.
• Identified a sufficient condition on resting time between attacks to ensure system resilience.