CE-PBFT: A high availability consensus algorithm for large-scale consortium blockchain
Jing Xiao,Tao Luo,Chaoqun Li,Jie Zhou,Zhigang Li
DOI: https://doi.org/10.1016/j.jksuci.2024.101957
IF: 9.006
2024-02-01
Journal of King Saud University - Computer and Information Sciences
Abstract:The consortium blockchain has been widely applied in various fields such as agricultural product traceability, supply chain management, and logistics transportation. As an indispensable component of a consortium blockchain, the consensus algorithm ensures the consistency and trustworthiness of each node in the network. However, existing consensus algorithms in large-scale consortium blockchain scenarios suffer from low system throughput and high latency due to the complexity of communication processes, rendering them impractical for real-world use. To address these issues, this paper proposes a novel consensus algorithm called credit evaluation-based practical Byzantine fault tolerance (CE-PBFT). This algorithm designs a new node credit evaluation model that considers node completion rate, consensus decay, and node behavior. It effectively measures and reflects the specific reliability status of nodes during system operation, thereby enhancing system reliability and security. Additionally, the paper introduces the innovative use of decision tree algorithms to analyze network node behavior and simplifies the existing consensus protocol. Nodes are categorized as excellent, good, ordinary, or poor based on the classification results, and non-Byzantine nodes are dynamically selected accordingly. This greatly improves the overall efficiency of the system. The performance of CE-PBFT is validated through experiments and compared with PBFT, G-PBFT, RBFT, WBFT and PPoR. Experimental results demonstrate that in large-scale consortium scenarios, CE-PBFT significantly improves system throughput, effectively reduces transaction latency and communication overhead, and outperforms the compared protocols.
computer science, information systems