Whole-Process Privacy-Preserving and Sybil-Resilient Consensus for Multiagent Networks.

Area of Science:

• Computer Science
• Network Security
• Distributed Systems

Background:

• Multiagent networks (MANs) face vulnerabilities from eavesdropping and Sybil attacks during information exchange.
• Existing methods for Sybil attacks often require node identification and mitigation, which can be complex and disruptive.

Purpose of the Study:

• To co-design a privacy-preserving and resilient consensus protocol for MANs.
• To address the challenges posed by eavesdropping and Sybil attacks in MANs.
• To develop a strategy that tolerates Sybil nodes without elimination.

Main Methods:

• Introduction of a new attack model for Sybil attacks in MANs.
• Design of a whole-process privacy-preserving mechanism to protect agent states.
• Implementation of a degree-based mean-subsequence-reduced (D-MSR) resilient strategy.

Main Results:

• The proposed privacy-preserving mechanism effectively protects initial and current agent states.
• The D-MSR strategy ensures the effectiveness of privacy protection and allows the network to contain Sybil nodes.
• Consensus is successfully reached among normal agents in the presence of Sybil nodes.

Conclusions:

• The co-designed protocol offers a robust solution for privacy and resilience in MANs under attack.
• The D-MSR strategy provides an effective alternative to Sybil node elimination for achieving consensus.
• Numerical simulations validate the effectiveness of the proposed privacy-preserving and resilient consensus protocol.