可验证随机函数的多主节点共识算法

Multi-primary-node consensus algorithm based on verifiable random functions

共识算法是区块链技术的核心,PBFT(practical byzantine fault tolerance)是广泛应用于区块链的一类共识算法.目前,现有的PBFT存在着主节点选取公平性问题、吞吐量较小以及通信时延较高等问题.为此,提出了一种基于可验证随机函数VRF(verifiable random function)的多主节点共识算法MVBFT(multi-primary-node verifiable-random-function byzantine fault tolerance).MVBFT引入了共识阶的概念,改进了区块链的区块结构,实现流水线式共识.其次,采用信誉模型,动态评估全网节点的信誉值.利用节点的信誉值和可验证随机函数,每个共识阶开始前随机选举出多个主节点,提升系统的安全性和可靠性.最后通过对PBFT算法的一致性协议进行改进,调整了网络拓扑结构,以减少通信量.仿真实验结果表明,与PBFT算法相比,所提出的MVBFT算法具有更低的通信开销和更高吞吐量,可以降低数据时延.

The consensus algorithm is the backbone of blockchain technology,PBFT(practical by zantinefaulttolerance)is a kind of consensus algorithm widely used in blockchain.At present,the existing PBFT has some problems,such as the fairness of primary node selection,low throughput and high communication delay.A multi-primary-node consensus algorithm MVBFT(multi-primarynode verified-random-function byzantine fault tolerance)based on verifiable random function VRF(verifiable random function)is proposed.First,MVBFT introduces the concept of consensus order,improves the block structure of blockchain,and realizes pipelined consensus.Second,a reputation model is used to dynamically evaluate the reputation value of the whole network nodes.Multiple primary nodes are randomly selected before the beginning of each consensus order by using the reputation values of the nodes and a verifiable random function,so that the security and reliability of the system are improved.Simulations conducted to evaluate the performance of MVBFT demonstrate its superiority over PBFT.The results indicate lower communication overhead,higher throughput,and reduced data delay.