A Study On Lightweight Data Integrity Auditing Schemes In Cloud Storage Environment

The cloud storage provides a pay-as-you-go storage model which supports flexible storage expansion.Therefore,more and more users are willing to store their data in cloud for saving cost.However,without the physical control of their outsourced data,the users worry about the security of their data.The integrity of the data is one of the users' most concerns.The data integrity auditing is an efficient method to solve the users' concerns.It's meaningful and challenging to improve the efficiency of the tag generation stage in public auditing.Certificated-based auditing not only simplifies the management of public key certificates in traditional public key auditing,but also avoids the key escrow problem in identity-based auditing.However,few scholars devote themselves to it.Real-time update of outsourced data is an important function of cloud storage,therefore dynamic data integrity auditing is more suitable for application.Non-repudiable data integrity auditing can protect the interests of cloud storage service providers form the dishonest users.So,it is very meaningful to improve the efficiency of data integrity auditing schemes;simplify the key management of data integrity auditing schemes;and design non-repudiable dynamic data integrity auditing schemes.Therefore,the thesis mainly researches several aspects mentioned above,and the main contributions of the thesis are summarized as follows.Firstly,since the user is likely to have information than the auditor in public auditing schemes,the verification phase is divided into private verification and public verification phases according to different situations.Then,a system model and a security model for public auditing with rapid private verification are proposed,which is described a new method of designing lightweight auditing schemes.In the system model,the user uses private verification phase to audit outsourced data promptly in most cases,and the auditor uses public verification phase to audit outsourced data only when disputes occurs or the user is not available to audit.Then,a scheme is proposed,and its security proof is presented in the random oracle model under the discrete logarithm assumption over a prime field.Experimental findings demonstrate that our auditing scheme only need tens of microseconds to generate data tags for every MB file outsourced.In other words,the efficiency of the our auditing scheme is almost as high as existing high-performing private auditing schemes,and it is higher in comparison to existing efficient public auditing schemes.Secondly,to simplify the key management is another direction of data integrity auditing schemes.Although like certificateless auditing schemes,certificate-based auditing schemes can simplify certificate management and mitigate key escrow concern,such schemes are not popular in the literature.Therefore,based on the model mentioned above,a model of the certificate-based public auditing with rapid private verification is proposed for improving efficiency and simplifying the key management.Then,a scheme is proposed based on asymmetric bilinear pairing,and its security is proven under the co-CDH assumption in the random oracle model.Then it is implemented by using type D curve of the pairing-based library.The implementation evaluation shows the proposed auditing scheme significantly reduces client's computation cost in the tag generation phase,particularly in comparison to several other competing schemes.For example,when the block size is 10 KB,the on-line computation cost of the proposed scheme in the tag generation phase takes only0.45 seconds,unlike other schemes that require at least 62.83 seconds under the same conditions.Furthermore,the analysis shows that when the number of blocks is fixed,the on-line computation cost of the proposed auditing scheme in the tag generation phase is constant despite varying file sizes.Finally,the real-time updating of outsourced data is an important function of cloud storage.For the scenes where clients are dishonest,a non-repudiable dynamic auditing model is proposed.Then,the notion of monotonic dynamic structure is proposed for non-repudiable dynamic auditing.Based on the Merkle hash tree and the map-version table,a monotonic dynamic structure(i.e.,index logic table)is proposed.Furthermore,a non-repudiable dynamic auditing scheme is proposed by using index logic tables and bilinear pairing.Analysis shows that the proposed scheme not only resists the hash value stored attack,the delete-insert attack,and the tampering cloud returned value attack,but also avoids the synchronization problem.Furthermore,the proposed scheme has lower computation cost and storage cost in the dynamic operation process.