Decentralized and Privacy Sensitive Data De-Duplication Framework for Convenient Big Data Management in Cloud Backup Systems

Abstract

The number of customers transferring information to cloud storage has grown significantly, with the rising prevalence of cloud computing. The rapidly rising data volume in the cloud, mostly on one side, is followed by a large replication of data. On the other hand, if there is a single duplicate copy of stored symmetrical information in the de-duplicate cloud backup the manipulation or lack of a single copy may cause untold failure. Thus, the deduplication of files and the auditing of credibility are extremely necessary and how they are achieved safely and effectively must be addressed in academic and commercial contexts urgently. In order to tune in this task by using application recognition, data similitude, and locality to simplify decentralized deduplication with two-tier internode and application deduction, we suggest a flexible direct decentralized symmetry deduplication architecture in a cloud scenario. It first distributes application logic to the contents of the directory through implementation-oriented steering to maintain a deployment location and also attributes the same kind of information to the cloud backup node with the storage node specificity by means of a hand printing-based network model to attain adequate global deduplication performance. We build up a new ownership mechanism during file deduplication to ensure continuity of tagging and symmetrical modeling and verify shared ownership. In addition, we plan an effective ownership policy maintenance plan. In order to introduce a probabilistic key process and reduce key storage capacity, a user-helped key is used for in-user block deduplication. Finally, the protection and efficiency audit demonstrate that the data integrity and accuracy of our system are ensured and symmetrically effective in the management of data ownership.