Traditional Byzantine consensus does not work in P2P network due to Sybilattack while the most prevalent Sybil-proof consensus at present still can'tresist adversary with dominant compute power. This paper proposed opiniondynamics based consensus for P2P network with trust relationships, consistingof the sky framework and the sky model. With the sky framework, opiniondynamics can be applied in P2P network for consensus which is Sybil-proofthrough trust relationships and emerges from local interactions of each nodewith its direct contacts without topology, global information or even sample ofthe network involved. The sky model has better performance of convergence thanexisting models including MR, voter and Sznajd, and its lower bound of faulttolerance performance is also analyzed and proved. Simulations show that ourapproach can tolerant failures by at least 13% random nodes or 2% topinfluential nodes while over 96% correct nodes still make correct decisionwithin 70 seconds on the SNAP Wikipedia who-votes-on-whom network for initialconfiguration of convergence>0.5 with reasonable latencies. Comparing tocompute power based consensus, our approach can resist any faulty or maliciousnodes by unfollowing them. To the best of our knowledge, it's the first work tobring opinion dynamics to P2P network for consensus.
展开▼
