The inefficiency of Consensus protocols is a significant impediment to blockchain and IoT convergence development.To solve the problems like inefficiency and poor dynamics of the Practical Byzantine Fault Tolerance(PB...The inefficiency of Consensus protocols is a significant impediment to blockchain and IoT convergence development.To solve the problems like inefficiency and poor dynamics of the Practical Byzantine Fault Tolerance(PBFT)in IoT scenarios,a hierarchical consensus protocol called DCBFT is proposed.Above all,we propose an improved k-sums algorithm to build a two-level consensus cluster,achieving an hierarchical management for IoT devices.Next,A scalable two-level consensus protocol is proposed,which uses a multi-primary node mechanism to solve the single-point-of-failure problem.In addition,a data synchronization process is introduced to ensure the consistency of block data after view changes.Finally,A dynamic reputation evaluation model is introduced to update the nodes’reputation values and complete the rotation of consensus nodes at the end of each consensus round.The experimental results show that DCBFT has a more robust dynamic and higher consensus efficiency.Moreover,After running for some time,the performance of DCBFT shows some improvement.展开更多
基金supported by the Science and Technology Plan Project of Quanzhou City,Fujian Province of China(2022C020R)the Science and Technology Plan Project of Fujian Province of China(2023H0012)。
文摘The inefficiency of Consensus protocols is a significant impediment to blockchain and IoT convergence development.To solve the problems like inefficiency and poor dynamics of the Practical Byzantine Fault Tolerance(PBFT)in IoT scenarios,a hierarchical consensus protocol called DCBFT is proposed.Above all,we propose an improved k-sums algorithm to build a two-level consensus cluster,achieving an hierarchical management for IoT devices.Next,A scalable two-level consensus protocol is proposed,which uses a multi-primary node mechanism to solve the single-point-of-failure problem.In addition,a data synchronization process is introduced to ensure the consistency of block data after view changes.Finally,A dynamic reputation evaluation model is introduced to update the nodes’reputation values and complete the rotation of consensus nodes at the end of each consensus round.The experimental results show that DCBFT has a more robust dynamic and higher consensus efficiency.Moreover,After running for some time,the performance of DCBFT shows some improvement.
