Research And Implementation Of Data Integrity Verification In Cloud Storage

The term cloud computing has recently become one of the most burning topics.With many prominent characteristics, it has been predicted as a potential drive thatcan lead to the next development of IT enterprise. When putting data andapplications remotely on the cloud, users are released from strenuous works ofmanaging hardware or software. In addition, users can access to their data orapplications anywhere and require as many resources as demand at a reasonable cost.Those merits make cloud computing outstanding from the traditional architecture, inwhich users maintain data on their own system, and adapted into practice morewidely. Cloud storage is one kind of service to prove that trend.It’s not like the traditional manner, in which data owner has full control overhis/her data, cloud storage systems provide users limited knowledge. Therefore,many security issues appear, including data leakage, data integrity etc. When usingcloud storage, users do not know exactly where their data are stored physically orwhether the storage medium is facing security threats, both malicious action andhardware failure. Thus, user’s data can be damaged randomly. This paper focuses onhow to verify data integrity in cloud storage. Our main contributions are as follows:1. Propose new data integrity verification scheme—DIV-I. The storageand communication overhead of DIV-I is similar to that of S-PDP scheme. However,the computation of DIV-I is less than that of S-PDP. In addition, DIV-I introducehow to update verification tags in case of dynamic data operations. Just like S-PDP,DIV-I also supports third party verification.2. Propose DIV-II scheme. Although the storage overhead of DIV-II isbigger than that of S-PDP, its computation time is proven a big advantage. Similar toDIV-I, DIV-I supports dynamic data operations and thirds party verification.3. Study sampling verification of two proposed schemes and optimizeverification plan. Probabilistic detection of a damaged file can be reached by computing part of the file to dramatically reduce the computation time of server andverifier. In this case, the computation is related to the required probability.