Research On The Dynamic Provable Data Possession Scheme Supporting Third Party Auditing In Cloud Storage

With the rapid development of cloud storage technology and the rapid growth in the amount of data,more and more people store their data in the cloud servers.However,although Cloud Storage brings many advantages,cloud users still are worried about the integrity of their data stored in the cloud servers because of the occurrence of the downtime accidents.Cloud users lose the physical control of their data after they upload their data to the cloud servers,so they can't detect the integrity of their data.Provable data possession can be used to detect the integrity of data.We call detecting the integrity of data as auditing the integrity of data.Now most existing provable data possession schemes do not support third party auditing or they are only applicable to static files.In this thesis a dynamic provable data possession scheme supporting third party auditing in Cloud Storage is presented,which allows cloud users to detect the integrity of their data without retrieving.In our scheme third party auditing is implemented by using algorithms that can be verified publicly.The third auditor can audit the integrity of data on behalf of the cloud users.The scheme prevents revealing user data information to the third auditor by choosing parameters randomly.The scheme is proved correct by carrying theoretical analysis.The probability of data integrity detection is discussed.The security of the scheme is proved by provable security theory.A secure scheme supporting batch auditing is presented,in which the third auditor can handle audit requests from different cloud storage users at the same time.It is proved that if a single auditing is not valid,batch auditing will fail with great probability.Batch auditing improvement strategies are given,and how to find invalid single auditing if batch auditing fails is discussed.This thesis presents a dynamic scheme supporting data modification,insertion,deletion.An authentication data structure called MB tree is designed,how to verify block tags and the comparison between the depth of MB tree and Merkle hash tree are given.At last,through the simulation experiments,the results show that the proposed scheme is as good as PDP(Provable Data Possession)scheme.The proposed secure batch auditing scheme does reduce auditing time,the improved method further reduces auditing time.The efficiency of the dynamic scheme is higher than others.