The Research And Implementation Of Multiple-replica Provable Data Possession In Cloud Storage

Cloud storage is a service provided by cloud service providers(CSPs).More and more organizations and individuals accept the service and outsource their data to a CSP to reduce local storage burden.In order to obtain both higher reliability and availability,users want to outsource multiple replicas.As users outsource more replicas,higher fees have to be paid to CSPs.Since the outsourced files are no longer under the data owners' direct control,CSPs can save storage space which yields higher benefits by saving fewer users' replicas.In addition,the integrity of the data stored in the cloud can be damaged,so it is important to study how to verify the integrity of multiple replicas stored in the cloud from the user's point of view.Provable data possession is the main method to verify the integrity of data stored in the cloud.Multi-replica provable data possession is developed and expanded on the basis of the traditional PDP scheme,it is a method to verify multiple replicas that are outsourced to the cloud.Multiple-replica provable data possession schemes such as an MB-PMDDP scheme usually use exponentiation operations to generate tags of file blocks.They then use homomorphic linear authenticators(HLAs)to generate aggregated tags for the blocks at the same indices in each replica,so they can verify the integrity of all replicas.The tag generation method requires heavy computation,and those schemes cannot verify single replica to identify the corrupted replicas.In this paper,based on the advantages and disadvantages of the existing multiple-replica provable data possession schemes,we propose a novel,flexible,multiplereplica provable data possession scheme named FMR-PDP.Our contributions are summarized as follows: 1)We propose a novel multi-replica provable data possession scheme,which can verify the integrity of the data stored in the cloud to make sure the CSP stores all replicas.2)We utilize vector dot product instead of expensive group exponentiation operations to generate tags for multi-replica,and only generate one tag for all blocks at the same indices in each replica.This method reduces the local computation of tag generation.3)It supports flexible data possession verification,which means that the data owner can check any number of replicas in a verification and find corrupted replicas.4)It also supports public verification.It permits the third party auditor to audit the integrity of data in cloud storage without private information.In this paper,after first formalizing the FMR-PDP scheme,we analyze the security of the scheme to ensure its accuracy and whether or not it can resist the replace attack,the forgery attack,and the replay attack.Then,we theoretically analyze the performance of our scheme via storage cost,communication cost,and computation cost.Finally,we implement the system based on this scheme.By comparing our FMR-PDP scheme and the MB-PMDDP scheme vis-à-vis the timing of tag generation,the timing of proof generation,and the timing of verification;we have proven our proposed FMR-PDP scheme has a performance advantage.