Data Security Storage And Deduplication Technolog

With the rapid development of computer networks,mobile Internet,and other information technologies,cloud storage is also thriving as an emerging network storage technology.Cloud storage technology stores data resources on cloud servers for users to access,effectively solving the problem of insufficient local storage capacity of users.However,as the scale of users expands,more and more duplicate data are stored on the cloud server side,bringing a huge waste of storage resources to the cloud server.Data deduplication technology is an effective method for cloud storage systems to improve storage efficiency,i.e.,cloud servers reduce data storage and transmission costs by storing unique copies of duplicate data.However,as users’ security awareness increases,they encrypted data before uploading them to cloud servers,which makes it difficult for cloud servers to identify the same data for data deduplication based on the ciphertext.How to improve the efficiency of data deduplication while better-protecting data privacy is a hot research issue in both academic and industrial fields today.In this paper,we address several problems in existing encrypted data deduplication schemes,focusing on the impact of user attributes on data deduplication techniques,and propose the following two solutions:1)We first propose the concept of clustering deviation to improve the DBSCAN algorithm,and propose a new algorithm called the DBSCAN With Tolerable Clustering Deviation(TCD-DBSCAN)algorithm,and apply it to the data deduplication process.We design a data deduplication scheme based on the TCD-DBSCAN algorithm.The scheme adopts the data popularity division strategy,considering user attributes and the popularity contribution value of data is determined by the similarity of attributes between the current uploader and previous uploaders,which reduces the risk of privacy leakage due to premature change of internal private data into popular data.Different encryption methods are adopted for data with different popularity to balance data security and computational overhead.The security analysis shows that the scheme has high data security.Finally,the simulation experiments verify the scheme’s usability and efficiency.2)We propose a data deduplication scheme with two-level authentication without the participation of trusted third parties.The data tags are constructed based on bilinear mapping to achieve Proof of Ownership(PoW).Considering the influence of user attributes on data deduplication,we propose the concept of Proof of Security(PoS)for the first time,to restrict key sharing among users and further guarantee the security of data.In this scheme,users adopt the ECDHE key negotiation algorithm for key sharing to ensure the semantic security of data.Meanwhile,the cloud server adopts Ciphertext-Policy Attribute-Based Encryption(CP-ABE)algorithm to re-encrypt the ciphertext uploaded by users to ensure the forward security of the system.Finally,the security analysis and simulation experiments verify that this scheme is secure and efficient.