Research On Secure Auditing In Cloud Storage

The explosive growth of Internet users and their data results in the rising storage cost of users' data,which increases the difficulty of data storage management.As a kind of "Infrastructure as a Service" form,cloud storage can provide a solution with low cost,high performance,virtualization and easy management.To make it convenient for users to store personal or enterprise data in the cloud,users can share data with each other anytime and anywhere.However,users have lost the direct control of data in the cloud storage.Therefore,the security management of data has become a negligible problem.Since the cloud server that provides storage service is not fully trusted,the security problem of cloud storage system is more complicated than ever.Users have reason to suspect whether their data are stored correctly in the cloud storage.Specifically,under cloud environment,the cloud server is affected by software,hardware,or human factors.Sometimes the cloud server may inevitably modify and even delete users' data.Besides,users delete their local copies after uploading data to the cloud server,which makes it impossible to verify the integrity of data directly.Therefore,it is meaningful to design a kind of scheme that supports integrity auditing of cloud storage data.Based on the existing data integrity auditing schemes,our dissertation mainly studies the problem of cloud storage data integrity auditing,and designs two new integrity auditing schemes.The main works lie in the following two aspects: 1.By using Bloom filters,we propose a privacy-preserving public auditing scheme which supports client-side deduplication.By eliminating user's private key in the response,our scheme achieves unconditional anonymity against third party auditor.In addition,since cloud server no longer needs to store duplicate data,our solution saves storage space of cloud server and reduces the bandwidth consumption of users in uploading files.The proposed scheme is proven to be uncheatable and anonymous under the variant of the BDH hardness assumption in the random oracle model.And security analysis indicates that our scheme is unforgeable during deduplication.2.We propose a public auditing scheme which supports client-side deduplication and strong key-exposure resilient.Third party auditor assists user in updating the key with its private key so that the key exposure in one time period does not affect the security of cloud storage auditing in other time periods.The security analysis shows that the proposed scheme is strong key-exposure resilient,confidentiality of file,detectability of file integrity,unforgeability of authentication tags and unforgeability of tokens.