Secure consensus for PDEs-based multiagent systems under DoS attacks via boundary control approach.

Area of Science:

• Control Theory
• Systems Engineering
• Applied Mathematics

Background:

• Multiagent systems (MASs) are crucial for distributed tasks but vulnerable to denial-of-service (DoS) attacks.
• Partial differential equations (PDEs) model complex distributed systems, posing unique consensus challenges.
• Existing consensus protocols often lack robustness against communication disruptions like DoS attacks.

Purpose of the Study:

• To develop secure consensus strategies for leader-following MASs modeled by PDEs under DoS attacks.
• To address challenges posed by communication channel paralysis and unmeasurable agent states.
• To introduce novel boundary control methods for enhanced system resilience.

Main Methods:

• Implementing buffer regions in communication channels to mitigate DoS attack effects.
• Utilizing observers to estimate unmeasurable states of leader and follower agents.
• Designing two boundary controllers: one using original boundary information, another using observed states.

Main Results:

• Achieved secure leader-following consensus in PDE-based MASs despite DoS attacks.
• Demonstrated the effectiveness of buffer regions in maintaining communication integrity.
• Validated the proposed controllers' ability to handle both measurable and unmeasurable states.
• This work represents the first use of buffers for PDE-based MASs under DoS attacks.

Conclusions:

• The proposed boundary control methods effectively ensure secure consensus in MASs under DoS attacks.
• The use of buffers and observers provides a robust solution for communication-constrained and state-unobservable systems.
• Boundary control offers a cost-effective approach by potentially reducing actuator requirements.