| As a typical computing model that develops rapidly and is promoted by the business, cloud computing leads to a revolutionary evolution in various fields, which has been widely recognized by various aspects for its unlimited expansion of computing resources, efficient elastic model and pay-on-demand philosophy. Public cloud, as the most promising one of the four kinds cloud computing deployments, faces a great obstacle, which is not the business capability but the security issues, especially the privacy information security problem for which users particularly concern.This paper designs the architecture of the public cloud platform privacy protection system that provides typical services to users. This architecture focuses on solving the following problem. First, the person other than the user himself, including the operator of the public cloud platform, should not be able to get the user’s privacy information in an unauthorized manner; Second, the user should be able to verify whether the services offered by the public cloud platform are consistent with the promises made by the cloud provider; Third, compared with to the loss caused by privacy leak, the costs increased by the privacy protection mechanism should be acceptable.Based on the analysis of typical service in the public cloud platform, such as scientific computing, this paper studies measures for protecting privacy. Especially, the paper design and implements the "computing service privacy protection" module. Based on the overall system architecture analysis, this paper presents the conceptual model of the computing services module. And for the linear equations and strictly convex quadratic programming algorithm that is commonly used in the scientific computation, this paper presents two kinds of sensitive data hidden technology, which are based on the Paillier homomorphic encryption technology and affine transform technology. The simulation results show the effectiveness of the implemented module. |
PDF Full Text Request