Research On Data Security Retrieval And Efficient Consensus Methods In Education Blockchain

With the deepening application of new technologies such as big data in the field of education,the sharing of educational data has become an inevitable trend that can promote the ecological development of the education industry.However,educational record data has its own peculiarities.On the one hand,educational data involves numerous participants and contains a large amount of personal privacy information,requiring privacy protection.On the other hand,it is necessary to ensure the reliability of the process of sharing educational data because this is crucial for users.Without reliability assurance,the purpose of querying the data,such as in the context of corporate recruitment,is lost.Existing technologies still have limitations in privacy security or correctness verification,making it difficult to achieve multiple guarantees of privacy security and data reliability.The traditional educational management model has directly led to the isolation of educational data,unable to meet the requirements of privacy,security,and quick exchange of rapidly growing educational data.In response to these demands,several solutions have been proposed: outsourcing a large amount of educational data to cloud servers to break the "data silo" and achieve convenient sharing,while reducing the storage pressure on educational institutions;protecting the security of educational data through encryption.However,there are still unresolved issues: 1)encrypted educational data suffers from the problem of losing data availability and operability,making it impossible to perform effective data search and analysis.In order to achieve efficient retrieval of encrypted data,some work has proposed outsourcing encrypted data indexing to cloud servers,but this introduces the new threat of malicious cloud servers.2)The commonly used consensus algorithm PBFT in educational blockchain has issues such as high latency and low consensus efficiency,which cannot meet the practical requirements of data retrieval and analysis in the education field,becoming one of the bottleneck technologies for educational data applications.To address these challenges,this thesis proposes a search service solution for encrypted educational data with multiple keywords based on blockchain and an improved low-latency PBFT consensus scheme based on reliability grouping.The main contributions are as follows:(1)For the issue of data trustworthiness,this thesis designs a trusted search method based on smart contracts.Firstly,searchable encryption technology is used to encrypt the data,and the encrypted data is stored on cloud servers to alleviate the storage pressure on the blockchain.At the same time,corresponding search smart contracts are designed on the chain to achieve trusted search of encrypted educational data.When conditions are met,the search smart contract is triggered and executed automatically,reducing external interference and the possibility of malicious attacks.The execution result of the contract is permanently stored in the blockchain ledger,and all nodes can verify it,ensuring the security and reliability of the search operation.For the issue of data security,this thesis designs an encrypted retrieval scheme with multiple keywords for educational data.Firstly,the keywords of encrypted educational documents are extracted and transformed into a vector structure using Bloom filters and hash functions.The index and query keywords are encrypted to ensure privacy preservation during the retrieval process.Furthermore,efficient multi-keyword matching retrieval can be achieved through vector inner product operations on the index and query keywords,achieving "availability without visibility" of the data.(2)This thesis designs a low-latency Byzantine consensus scheme DPBFT suitable for education blockchain,which overcomes the issues of high latency,high communication overhead,and poor robustness in the PBFT consensus algorithm in education blockchain.It selects reliable nodes for consensus based on a reliability calculation algorithm,avoiding interference from malicious nodes in the consensus process.At the same time,it selects backup primary nodes within the reliable node group to solve the problem of primary node failure in PBFT.Through experiments,the DPBFT algorithm has been demonstrated to have advantages such as low latency,high reliability,and high efficiency.