Research On Public Auditing For Shared Data And Assured Deletion In Cloud Storage

Recently,as an important service of cloud computing,cloud storage develops rapidly with more and more data moved into the cloud.In the paradigm of cloud storage,the users outsource their data hosting to the cloud provider and access,handle and share their outsourced data through Internet.However,although this new paradigm of data hosting brings much convenience and cost savings for the users,it also brings about many new security challenges.Considering the whole life cycle of data in the cloud,the users first need to ensure that the outsourced data is correctly stored in the cloud,preventing the integrity of outsourced data from destroying.Furthermore,the users need to ensure that the data in the cloud will not resurface after deletion.To tackle these challenges,the main contributions of this thesis are as follows:1)An efficient and secure integrity auditing scheme for shared data in the cloud.The existing schemes assume that there is no collusion between the revoked users and the cloud.However,collusion attacks of this kind will not only allow the revoked users to generate valid signatures,but also disclose secret keys of other users.Moreover,the computation overhead and communication cost of the existing schemes are linear to the group size.To solve these problems,an efficient and secure integrity auditing scheme for shared data in the cloud has been proposed in this thesis.By leveraging the concept of threshold secret sharing,the proposed scheme splits the re-signing process into a number of sub-processes and deploys them to several different proxies,which makes the collusion attacks practically infeasible.Besides,the utilization of our improved polynomial-based authentication tags makes the auditing process secure and efficient.Numerical analysis and experiment results demonstrate that the proposed is collusion-resistant and efficient.2)An secure assured deletion scheme for cloud storage based on AONT.To achieve assured deletion in the cloud,the existing schemes choose to encrypt data before outsourcing and destroy encryption keys when deleting.However,the existing schemes only destroy the encryption keys and leave the ciphertext intact in the cloud,which may lead to privacy disclosure in the future.To solve these problems,this thesis propose an secure assured deletion scheme for cloud storage based on AONT.The proposed scheme combines the AONT and data encryption to transform the original data into strong nonseparable ciphertext,which means that it is infeasible to determine any original message block before decrypting all the ciphertext.When deleting,a part of the ciphertext and the encryption key are destroyed simultaneously,which violates the integrity of the ciphertext and increases the validity of assured deletion.