Cryptography Technology Of Secure And Reliable Cloud Storage

Cloud storage has been a new service with high scalability,high reliability,low price and other characteristics.Users store their data in data centers provided by cloud providers by ”on-demand self-service”,however users can not manage their data as local.What most matters for users are security,integerity,and availability,therefore,how to guarantee security,integerity,and availability of data motivates many concentrations in industry and academic communities.General data verification methods have to download the data from cloud,which causes huge burden to network,consequently,those methods are not common accepted.Fortunately,some new verification algorithms based on public-key cryptography e.g.RSA are applied for guaranteeing confidentiality of verification process and low networking cost.However,RSA involves a large number of exponential operation which lead to high computational overhead.Outsourced data is dynamic,users can add,delete and update data,which posts challenges to data integrity verification.To overcome those challenges,a variety of dynamic verification schemes had been proposed,most of which are based on Merkle Hash Tree.With the development of quantum computer,cryptographic schemes based on traditional mathematics problem are no longer safe,such as RSA which was built upon large integer factoring problem.Therefore,higher requirements are proposed to data verification:(1)cryptography method can resist quantum attack;(2)public authentication supported;(3)privacy protection;(4)dynamic authentication;(5)high efficiency.This thesis takes the cloud storage integrity verification as the research object,and two aspects of the work is completed as follows:1.This thesis proposed a new scheme,named cloud storage data dynamic integrity and availability verifications with trusted third party authentication,by introducing lattice linear homomorphic signature(LHS)and Merkle Hash Tree(MHT),which meets five requirements.The scheme introduces trusted third party authentication in order to support public audit,computes the value of the root node by the MHT definition,and then,the root node is used to determine the integrity of the cloud data blocks.The basic algorithm is implemented using Python programming language.2.This thesis describe and analyze the file operation and data verification mechanism of HDFS clusters,and present the disadvantage of verification —cannot resist quantum attack.The verification scheme we proposed has high requirements for users computing power,in order to reduce the burden,We present a server between users and HDFS,explain document management and preprocessing of verification process particularly.