The K-DPBFT Algorithm based on Byzantine Fault Tolerance Algorithm
Yan Li, Jiulun Fan, Gang Han · 2024
The PBFT algorithm, a classic Byzantine Fault Tolerance consensus mechanism, is designed to mitigate node failures and malicious behavior within distributed systems. However, its effectiveness is hindered when confronted with large-scale networks and heightened latency. In response to these challenges, we proposed the K-DPBFT algorithm, a novel approach based on Byzantine Fault Tolerance principles, aimed at overcoming the aforementioned limitations.The dataset is initially partitioned through multiple iterations, merging clustering results from different partitions. Subsequently, the data is clustered using the DBSCAN algorithm. Finally, two rounds of the PBFT consensus algorithm are conducted to achieve node consensus, aiming to strike a balance between efficiency and performance. Experimental results demonstrate that, in a consensus process involving a significant number of participating nodes, the K-DPBFT algorithm, while possessing Byzantine Fault Tolerance capability, reduces communication interactions by over tenfold compared to the traditional PBFT algorithm. Moreover, the consensus duration is shortened by more than 20%, significantly enhancing consensus efficiency.