Research And Application Of PBFT Consensus Mechanism Based On Trust Model

In recent years,blockchain technology has received a lot of attention,because blockchain technology can not only serve as a trusted third-party platform,but also strengthen data tracking and traceability management.The consensus mechanism used for blockchain nodes to reach agreement is an important content of blockchain technology research.Among them,the practical Byzantine fault tolerance(PBFT)consensus mechanism solves the problem of low efficiency of the Byzantine fault tolerance(BFT)algorithm.PBFT makes the Byzantine fault-tolerant algorithm have practical application feasibility,thus becoming a consensus mechanism that is widely used in the process of technical engineering practice.However,the consensus mechanism still has problems such as large consensus delay caused by random node selection and insufficient node control.This paper studies the PBFT-based blockchain technology and proposes a trust model-based PBFT consensus mechanism.This article applies TMPBFT to the design and implementation of rice traceability systems.The main contents of this article are as follows:Aiming at the problem that the random selection of nodes in the consensus mechanism leads to a large consensus delay and insufficient node control,a trust modelbased PBFT consensus mechanism is proposed.The consensus mechanism includes communication and processing power preference strategy and node classification processing method based on trust model.Among them,(1)communication and processing power priority selection strategy used to solve the problem of large consensus delay caused by node random selection.Combining communication time and processing time,each node establishes and maintains a node selection table.The node with short communication time and short processing time is given priority in the table.And when sending a message,each node selects the node to send the message according to the order in the table.This strategy reduces the idle waiting time of good nodes and promotes the reduction of consensus delay.(2)Node classification processing method based on trust model is used to solve the problem of insufficient node control.This method is based on the communication and processing capacity priority selection strategy.This method improves the P2 P trust model based on contextual factors.This method comprehensively judges the node situation by multi-factors such as the frequency of node participation and the importance of transactions.The method gives different trust states according to its situation,and classifies the nodes according to different trust states.This method allows trusted nodes to act as master nodes and restrict malicious nodes from participating in consensus,improving the node's fault tolerance of the PBFT consensus mechanism.In addition,this method improves system performance by increasing the voting weight of trusted nodes,reducing the voting weight of failed nodes and malicious nodes,reducing the number of nodes needed to complete consensus..Combined with TMPBFT,design and implement rice traceability system.This article firstly analyzes the requirements of the system,and then conducts detailed design of the system architecture,business processes,functional modules,and databases.On this basis,the Hyperledger Fabric framework is used to implement the rice traceability system,and the system performance is tested to ensure that the system can meet the needs of use.