Research On Consortium Chain Consensus Algorithm Based On BP Neural Network And Credit Mechanism

As a new generation of information technology,blockchain has become an important part of the digital economy and has attracted extensive attention from the academic and industrial circles.The Consortium Chain has attracted extensive attention due to its access node identity controlled,decentralized storage and other characteristics.It has become the preferred framework for the application of blockchain,and has been widely used in digital finance,product traceability,identity authentication and other fields.As a key technology in the consortium chain,the consensus algorithm directly affects system performance.A good consensus algorithm can not only reduce the consensus time but also improve the security of the system,which has become the focus of research by experts and scholars in the blockchain field.Based on the overview of the current mainstream consensus algorithms,this paper conducts in-depth research on the problems of low throughput,poor security and scalability of the existing consensus algorithms of the consortium chain.The main work completed is as follows:(1)Comparison of consensus algorithms in the consortium chain.Summarize the existing mainstream consortium chain consensus algorithms,analyze the advantages and disadvantages of various algorithms,and point out the main problems of existing consortium chain consensus algorithms.(2)Theoretical innovation.Aiming at the limitations of the existing consensus algorithm of consortium chain in scalability,concurrency and security,this paper proposes a consensus algorithm of consortium chain based on segmented DAG and BP neural network.First,a data storage structure based on segmented directed acyclic graph is proposed to improve the scalability,throughput and fine-grainedness of transaction data.And designed a Map Reducebased anti-double spending mechanism to ensure that the transaction data is globally unique and orderly.Secondly,a new node reputation evaluation mechanism based on BP neural network is designed to improve the accuracy of node reputation evaluation and reduce the possibility of Byzantine nodes.Finally,according to the node credit value,the accounting nodes are quickly selected to improve the efficiency of the algorithm.Experimental results and security analysis show that the algorithm has advantages in throughput,and has higher security and scalability.(3)Application innovation.Based on the proposed consensus algorithm of the consortium chain,this paper constructs a vaccination information management system based on the blockchain.The system takes fabric as the prototype framework and transplants the consensus algorithm proposed in this paper.In terms of data storage,according to the data characteristics of vaccination information,this paper adopts the method of storing data on the chain,dividing nodes into light and heavy nodes,which store all data,and light nodes store the offset between the hash of data and the file,effectively reducing the storage pressure of data on the chain.In terms of data encryption and sharing,the Chinese Remainder Theorem is used to construct a congruence equation set to segment and reunite data,which can effectively ensure the security of data,and realize the reproduction of data through their own keys and the data after segmentation in the event of an emergency that they cannot authorize.The effectiveness of the proposed mechanism is demonstrated through simulation experiments and practical applications,which provides a reference for the innovative development and application of blockchain.