Research On Dynamic Data Provenance And Detection Algorithm Based On Blockchain

Data opening and sharing is an important means to connect the islands of information and activate digital assets.However,the subsequent problems that privacy leaks and illegal transactions were particularly prominent,which brought unprecedented challenges to the security protection of big data and the protection of digital rights.Data provenance is to trace the original data and reproduce its evolution process.Through tracing,people can understand the change process of data life cycle to ensure data security and credibility.The blockchain is tamper-proof and can ensure the authenticity of data.Digital watermark has the special property of hiding information and it can complete the anti-counterfeiting of data.In view of this,based on blockchain technology and digital watermarking technology,this paper conducts research on data provenance to provide security guarantee for the opening and sharing of big data.The specific work is as follows:(1)Aiming at the problem of data sharing and privacy protection of blockchain data tracing algorithm,this thesis proposes a blockchain data tracing algorithm based on attribute encryption.First,a strategy update algorithm suitable for blockchain is designed based on the CP-ABE scheme proposed by Waters to complete the dynamic protection of transaction privacy.Secondly,the traceability chain structure is designed based on the strategy update algorithm to complete the dynamic update of the block content visibility and realize the dynamic sharing of traceability information.Security and experimental simulation analysis show that the proposed algorithm can realize the dynamic sharing of traceability information while completing the protection of transaction privacy.(2)Aiming at the problem of low access efficiency of blockchain traceability information,this thesis proposes a blockchain data provenance scheme based on grouping consensus and B-M tree.First,a group consensus mechanism is constructed based on the PBFT consensus algorithm to improve the speed of transaction confirmation and complete the rapid write of information.Then,the B-M transaction tree is designed basedon the AVL tree and the Merkel tree to ensure the traceability information is effectively queried under the premise that the transaction information is not tampered.The correctness,security and experimental simulation analysis show that this solution has certain advantages over the traditional blockchain tracing scheme based on Ethereum in terms of traceability information on-chain and query.(3)Aiming at the problems of small amount of embedded information,weak robustness and poor concealment in text watermarking,this thesis proposes a text watermarking algorithm based on DNA coding and scrambling diffusion.First,the watermark information encoding is designed based on the invisible characteristics of part of the ASCII code to realize the hidden embedding of a large amount of watermark information.Then,the row and column diffusion algorithm is designed based on the random scrambling thought to achieve the robust blind extraction of watermark information.Experimental analysis shows that this algorithm can hide a large amount of watermark information into the text carrier,and its robustness has obvious advantages compared with other algorithms.