Secure Event-Triggered Consensus Control of Linear Multiagent Systems Subject to Sequential Scaling Attacks.

Area of Science:

• Control Systems Engineering
• Cybersecurity for Multiagent Systems
• Networked Systems Theory

Background:

• Linear multiagent systems are vulnerable to deception attacks impacting secure consensus.
• Existing defenses often use probabilistic models, lacking specific attack parameterization.
• Event-triggered control offers communication efficiency but needs robust security against sophisticated attacks.

Purpose of the Study:

• To develop a resilient event-triggered control framework for linear multiagent systems.
• To defend against a novel sequential scaling deception attack by defining its properties.
• To propose and analyze static and dynamic event-triggered protocols for enhanced security and communication efficiency.

Main Methods:

• Investigated static event-triggered control, deriving consensus conditions constrained by attack duration and frequency.
• Introduced a dynamic event-triggered scheme with a state-based auxiliary variable.
• Analyzed consensus criteria involving triggering parameters, attack constraints, and system matrices, proving Zeno behavior exclusion.

Main Results:

• Sufficient consensus conditions established for static event-triggered control, defining limits on attack parameters.
• Novel consensus criteria derived for dynamic event-triggered control, integrating system, attack, and triggering parameters.
• Demonstrated exclusion of Zeno behavior in the dynamic event-triggered scheme.

Conclusions:

• The proposed static and dynamic event-triggered control mechanisms effectively defend against sequential scaling deception attacks.
• The framework establishes a clear relationship between attack characteristics (duration, frequency) and event-triggered parameters.
• Validated effectiveness through simulations, highlighting the impact of scaling factors and attack properties on system performance.