基于积分机制的实用拜占庭容错算法优化  

Study on Optimization of Practical Byzantine Fault-Tolerant Algorithm Based on Cumulative Scoring Mechanism

作　　者：郭天元 于瓅[1] 黄金炜 GUO Tianyuan;YU Li;HUANG Jinwei(College of Computer Science and Enginering,Anhui University of Science and Technology,Huainan Anhui 232001,China)

机构地区：[1]安徽理工大学计算机科学与工程学院,安徽淮南232001

出　　处：《重庆科技学院学报（自然科学版）》2023年第5期51-55,共5页Journal of Chongqing University of Science and Technology：Natural Sciences Edition

基　　金：2021年安徽省重点研究与开发计划项目“基于多源数据融合的交通事件主动防控技术研究与示范应用”(202104D07020010)。

摘　　要：实用拜占庭容错(PBFT)算法是一种广泛应用于联盟链的共识算法,但也存在主节点选举随意、通信复杂度高和可扩展性差等问题。为此,对PBFT算法加以改进,提出一种结合积分机制的PBFT优化算法——OPBFT算法。首先,结合积分机制,基于节点在共识过程中的表现对其进行评分;然后,根据节点积分随机选择主节点;最后,根据联盟链的环境对一致性协议进行优化,减少了一次全网广播的过程。实验结果表明,相较于PBFT算法,OPBFT算法在共识时延和吞吐量等方面有着更好的表现。Practical byzantine fault tolerance(PBFT)is a consensus algorithm which is widely used in alliance chain,but it also has some problems such as arbitrary primary node selection,high communication complexity and poor scalability.Therefore,the PBFT algorithm is improved and a PBFT optimization algorithm—OPBFT algorithm which combines cumulative scoring is proposed.Firstly,combined with the scoring mechanism,nodes are scored based on their performance in the consensus process.Then,the main node is selected randomly according to the score of the node.Finally,the consistency protocol is optimized based on the environment of the alliance chain,reducing the need for a whole network broadcast.The experimental results show that compared with PBFT algorithm,OPBFT algorithm has better performance in terms of consensus delay and throughput.

关 键 词：区块链 拜占庭容错 共识算法 积分机制 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]