Improvement of Practical Byzantine Fault Tolerant Consensus Algorithm Based on Reputation Evaluation and Threshold Signature
Xuefeng Long, Xiaofang Wang, Pengfei Lu, Jiaquan Song, Zitong Shao · 2024
Blockchain ensures data consistency and security through consensus algorithms. The PBFT is a consensus mechanism frequently applied in consortium blockchains. However, PBFT encounters performance bottlenecks when handling a substantial number of nodes, thereby significantly constraining its applicability in blockchain systems. This paper proposes an enhanced PBFT algorithm called RT-PBFT. The algorithm designs a reputation model to assess the credibility of nodes, then groups nodes based on the evaluation results, and subsequently divides the consensus process into intra-group and inter-group consensus, thus improving overall consensus efficiency. Additionally, a threshold signature mechanism is incorporated to optimize intra-group consensus, effectively reducing the communication overhead among nodes. The performance of RT-PBFT algorithm is verified through comparative experiments with traditional PBFT and G-PBFT algorithm. The experimental results demonstrate that the RT-PBFT algorithm reduces transaction latency and enhances throughput.