摘要
基于互联网的快速发展及隐私保护的迫切需求,区块链技术在信息共享及信任领域愈发普及与发展,而共识算法作为区块链的底层关键技术,其效率决定着区块链应用的效果.自PBFT算法提出后,已出现了不少基于该算法的改进方案,但网络实时性差及多节点场景下的各种PBFT算法优化方案效果仍不尽人意.本文基于Raft算法进行分层,通过在领导层共识过程中引入BLS门限签名,提出一种低延迟、高鲁棒性的拜占庭容错共识算法,并提出一种基于信誉值的动态分组策略,避免了在同一组中出现多数拜占庭节点的情况,提升了方案的安全性,进一步保证了算法的活性.仿真环境实验测试结果表明,在网络性能差及节点更多的环境下,该算法比PBFT算法性能更好且具有更高的扩展性.
Based on the rapid development of the internet and the growing urgent demand for privacy protection,blockchain technology has become increasingly more popular and developed in the field of information sharing and trust,and the consensus algorithm,as the underlying key technology of blockchain,its efficiency determines the quality of blockchain application.Since the proposal the PBFT algorithm,there have been many improvement schemes based on it,but the effect of various PBFT optimization schemes in poor network real-time and multi-node scenarios is still unsatisfactory.Layering based on Raft algorithm,this paper proposes a low delay and high robustness threshold signature Raft Byzantine fault tolerance(TRBFT)by introducing BLS threshold signature into the leadership consensus process.Meanwhile,a dynamicgrouping strategy based on reputation value is proposed to avoid the situation that most Byzantine nodes appear in the same group,which improves the security of the scheme and further ensures the liveness of the algorithm.Finally,the experimental test results of the simulation environment show that the TRBFT has a better performance and a higher scalability compared with the original PBFT in the environment of poor network performance and more nodes.
作者
刘金涛
沈丽敏
Liu Jintao;Shen Limin(School of Computer and Electronic Information,Nanjing Normal University,Nanjing 210023,China)
出处
《南京师范大学学报（工程技术版）》
CAS
2023年第4期10-18,共9页
Journal of Nanjing Normal University(Engineering and Technology Edition)
基金
国家自然科学基金青年基金项目(61802195)。
关键词
区块链
共识算法
门限签名
信誉分组
RAFT
blockchain
consensus algorithm
threshold signature
reputation grouping
Raft
