Research On Security Of Outsourcing Data Based On Cloud Platform

As the information is developing with high speed, computer network hasbecome the most important way to share information, and the emergence of cloudmakes the information share more convenient, cloud platform provide softwareand hardware, data resources and other services to the end users, users need touse the client access to the Internet only, it’s convenient to use a cloud platform toprovide resources. With the development of cloud computing, cloud securitydraw attention of public as the development of cloud, especially some problemscaused by security, such as low efficiency, large overhead, data leakage, etc.This article mainly refers to the following three aspects to study the securityof cloud, the security model of cloud system, mechanism of security index andmethods of key management.The article first introduces the data model of cloud which is the based onoutsourcing system and propose the problem that exist in the model, then we putforward the over-encryption model, data block should be encrypted beforetransforming to the cloud, when end users apply for data resources, the encrypteddata should be encrypted again before the cloud return to end users. This paperanalyzes the necessity of over-encryption, then we improve the over-encryptionand analyze the efficiency and feasibility of the over-encryption. The informationis storage in the cloud in the form of encryption, end users need to decrypt theencrypted data to obtain real information, which makes management of keybecomes particularly important, this paper puts forward the method ofmanagement tree based on the derivation, user could derive all the key that heshould know from existing keys and some public information. It allows users tomaintain letter keys so as to reduce the burden of client.The problem of index security should not be ignored. Security index mechanism is based on the data block, we divide the data according to barrel andestablish B+in the barrel in private cloud in order to improve the efficiency ofretrieval, and then blur the keywords of user retrieval in the end user terminal inorder to protect the privacy of authorized user. Data shuffle mechanism, indexupdate mechanism and cache replacement strategy are put forward in this articleto prevent the cloud from digging the user retrieval information maximized. Theexperiment is used to analysis the overhead and feasibility of keywords shuffling,index establish, index update, data reconstruction strategies, etc.