| The development of the network era has produced a large amount of data,and data has become the fifth important basic factor of production after material,human and other factors of production.In reality,the vast majority of online platforms are data-driven,but there is not enough data in their own ecosystems.To truly realize the full potential of big data,data must be flowing and turned into assets.There are many problems in data asset circulation transactions,such as the typical problems of massive data storage scalability,real-time computing,and transaction reliability,traceability,fairness and transparency.In this study,a blockchain-based framework for cloud side collaboration is designed by using the idea of industrial cloud side collaboration and combining with the blockchain technology widely used in existing studies to effectively solve many problems.For example,the problem of extensive data sources,a large number of computing and storage resources can not meet the problem,the problem of real-time pricing and data transaction rights and credibility.After determining the mode of using blockchain and cloud-side collaboration,the key technologies in the architecture are studied.On the premise that the default transmission performance does not affect the performance of the overall framework,the key point limiting the overall performance of the cloud-side collaboration framework is the cloud,which is mainly limited by the performance of the blockchain.The most resource-consuming step in blockchain is its consensus process.Practical Byzantine Fault Tolerance(PBFT)algorithm is commonly used in the consensus process.However,PBFT algorithm has the problem of excessive resource consumption.Therefore,this paper changes the consensus mechanism based on PBFT algorithm to RePBFT algorithm.Experimental results show that the improvement is effective and significantly improves the performance of blockchain.In the framework design stage,the overall architecture is firstly designed in detail,the behaviors of the three main participants(data holders,data buyers and platform side)are clearly defined,and the interaction modes of edge end,cloud and blockchain are determined.In the design of the block chain system,the Fabric platform is selected as the basis for the development of the system by quantifying the performance of Fabric and Ethereum.After that,the structure of the block chain system is analyzed and designed.Finally,the operation logic of the smart contract is sorted out and designed.In the implementation stage,the cloud chooses to build Hadoop cluster and block chain for implementation,uses Fabric-Java-SDK tool for development at the edge,and uses MySQL database to realize smart contract version control and information storage.After all the steps are completed,the system is tested in the test environment for functional testing and performance testing,and the test results show that the system can achieve the expected effect. |