EIoT-PBFT: A multi-stage consensus algorithm for IoT edge computing based on PBFT

With the introduction of edge computing into the field of Internet of Things (IoT), the Cognitive Internet of Things (CIoT) has emerged as the next-generation solution for trust and intelligent reasoning in the IoT. That also puts blockchain, with its unique consensus mechanism, transparency and trustworthiness, on the stage of IoT applications. At present, not much research is focused on blockchain's application in CIoT, whose development is to a large extent restricted by the inefficiency of the consensus algorithm. Considering the characteristics of CIoT, a multi-stage consensus algorithm of EIoT-PBFT is proposed on the basis of PBFT algorithm, which includes the Grouping stage, Scoring stage and Consensus reaching stage. EIoT-PBFT meets the IoT edge computing setup by adopting a two-phase improved PBFT algorithm and a scoring mechanism based on both location and reputation, thus achieving a great increase in consensus efficiency. Evaluation results show that EIoT-PBFT takes 36.4% less time than PBFT for a single consensus, and the performance remains stable over the 2500 node configurations we set up. Moreover, at a scale of 1000 nodes, the number of edge nodes to be configured to reduce the number of system communications by 90% compared to the PBFT algorithm is only 5, making blockchain more customized for CIoT settings.

Automated summary

This article discusses the research progress on the application of blockchain in Cognitive Internet of Things (CIoT) with edge computing, as well as the efficiency issue of existing consensus algorithms. Based on this, a multi-stage consensus algorithm of EIoT-PBFT, which improves consensus efficiency by refining the algorithm and introducing a scoring mechanism, is proposed.