The Emergence of Consensus from P2P Network with Trust Relationships

The decentralized nature of P2P network increases robustness because it removes single points of failure, however, traditional Byzantine consensus does not work in P2P network due to Sybil attack while existing Sybil-proof consensus based on compute power can't resist adversary with dominant compute power. We proposed an opinion dynamics model for almost-everywhere consensus, the model is Sybil-proof through trust relationships, and consensus emerges from local interactions of each node with its direct contacts without topology or global information of the network involved. We also proved the lower bound of the model's fault tolerance performance. Our model enables disarming faulty or potentially malicious nodes by unfollowing them, and to the best of our knowledge, it's the first work of consensus over P2P network by opinion dynamics.