Research And Application Of Consensus Algorithm Based On Fault-tolerant Mechanism In The Internet Of Things

The advantages of blockchain are mainly manifested in two aspects,namely decentralization and immutability of data,which enables the technology to be applied to many fields including the Internet of Things.However,since the various functions of the current blockchain system are not yet fully perfected,there is still a big gap between the performance requirements of the Internet of Things application scenarios.The performance of the consensus algorithm determines the performance such as delay to a large extent.Consensus algorithm is at the core of blockchain technology.It can provide support for blockchain nodes through large-scale collaboration between nodes to deal with trustless and decentralized network environments,so as to give full play to its performance advantages and further Change the system throughput,transaction confirmation delay,application scenarios and other aspects of blockchain technology.In view of the problems of PBFT algorithm such as random selection of master nodes and infringement of network resources,this paper proposes two improved PBFT algorithms based on Byzantine fault tolerance mechanism for two common Io T environments.The specific content mainly includes:(1)In the Internet of Things environment,there are big differences in performance and security among nodes.However,the traditional PBFT algorithm has great arbitrariness in the selection of the master node and does not deal with illegal nodes,which in turn leads to frequent view switching and reduces the consensus efficiency.Therefore,a scoring and sorting mechanism is added when the algorithm is improved,and the master node is selected rationally through the scoring and sorting method to avoid the above problems.Through comparison and simulation,the final conclusion is:the improved algorithm can effectively avoid malicious nodes in the selection of master nodes,and tends to high-quality nodes infinitely,greatly reducing the number of view switching.(2)In a large-scale Io T environment,the number of nodes is large and their functions are complex,and there is often a demand for dynamic changes.The traditional PBFT network has the characteristics of a fixed number of nodes,and it is necessary to initialize the consensus network to realize the increase or decrease of nodes;in terms of network bandwidth,the bandwidth increases at a polynomial level with the increase of the number of nodes,and the efficiency is serious when there are more nodes.decline.Therefore,this article introduces a threshold signature to replace the traditional PBFT algorithm,which requires two-to-two communication,to reduce the communication overhead.At the same time,a solution to support dynamic changes of nodes is proposed,a weak center routing table is designed,a node state table is formulated,a message process is set for node entry and exit,and malicious nodes are included in the blacklist.The simulation results show that compared with the traditional PBFT algorithm,the improved algorithm has achieved increased functions,can dynamically change each node,and the improvement in performance indicators such as system throughput and transaction confirmation time is also in line with expectations.