基于RTM信任模型的DPoS共识机制改进研究  

Improvement of DPoS Consensus Mechanism Based on RTM Trust Model

作　　者：高玮军[1] 张小芳 GAO Wei-jun;ZHANG Xiao-fang(School of Computer and Communication,Lanzhou University of Technology,Lanzhou 730050,China)

机构地区：[1]兰州理工大学计算机与通信学院,甘肃兰州730050

出　　处：《计算机技术与发展》2024年第10期93-99,共7页Computer Technology and Development

基　　金：国家自然科学基金(61762059)。

摘　　要：为了解决传统DPoS共识算法中存在的恶意行为“摇摆节点”问题和因权益分配不均所造成的“财富集中”问题,提出了一种基于RTM信任模型改进的R-DPoS共识算法。首先,该算法根据节点间的信息传播方向与交易关系对网络节点进行集合分类,在考虑了时间衰减因子等影响因素的条件下分别计算了各节点的直接信任值、推荐信任值以及全局信任值;其次,引入了节点摇摆度值概念,根据计算得到摇摆值的大小,将相邻共识轮次前后该值的变化情况进行比较,再对摇摆节点做出不同程度下的共识轮次惩罚限制措施;最后,在见证人节点投票过程中,增加了平衡反对票环节,再利用前期得到的综合信誉值,构建了一套新的记账节点选举评估准则,根据每轮计算后得到的节点评估值大小来动态地选择见证人节点,并完成一轮共识过程。仿真实验结果表明:改进后的共识算法相较原始DPoS共识算法、基于奖励机制和信用机制的改进共识算法而言,节点选举成为见证人节点的积极性提高了24%、18%,公平性增加了15%、10%,节点作恶的概率降低了9%、15%。In order to solve the problem of"swing node"of malicious behavior and"wealth concentration"caused by unequal distribution of rights and interests in traditional DPoS consensus algorithm,an improved R-DPoS consensus algorithm based on RTM trust model is proposed.Firstly,the algorithm classifies the network nodes according to the information propagation direction and transaction relationship between the nodes,and calculates the direct trust value,recommended trust value and global trust value of each node respectively under the condition of considering the influence factors such as time attenuation factor.Secondly,the concept of node wobble value is introduced.According to the calculated wobble value,the change of the value before and after the adjacent consensus rounds is compared,and then the consensus rounds penalty restriction measures are made for the wobble nodes in different degrees.Finally,in the process of witness node voting,the balance of negative votes is added,and a new set of election evaluation criteria for accounting nodes is constructed by using the comprehensive credit value obtained in the previous stage.Witness nodes are dynamically selected according to the value of node evaluation obtained after each round of calculation,and a consensus process is completed.The simulation results show that compared with the original DPoS consensus algorithm and the improved consensus algorithm based on reward mechanism and credit mechanism,the initiative of node election to become a witness node is increased by 24%and 18%,the fairness is increased by 15%and 10%,and the probability of node evil is reduced by 9%and 15%.

关 键 词：区块链 信任模型 DPoS共识算法 摇摆节点 信誉值 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]