A SEMANTIC BASED POLICY MANAGEMENT FRAMEWORK FOR CLOUD COMPUTING ENVIRONMENTS

摘要

Cloud computing paradigm has gained tremendous momentum and generated intensive interest.udAlthough security issues are delaying its fast adoption, cloud computing is an unstoppable force and we need to provide security mechanisms to ensure its secure adoption.ududIn this dissertation, we mainly focus on issues related to policy management and access control in the cloud.udCurrently, users have to use diverse access control mechanisms to protect their data when stored on the cloud service providers (CSPs).udAccess control policies may be specified in different policy languages and heterogeneity of access policies pose significant problems.An ideal policy management system should be able to work with all data regardless of where they are stored.udSemantic Web technologies when used for policy management, can help address the crucial issues of interoperability of heterogeneous CSPs.ududIn this dissertation, we propose a semantic based policy management framework for cloud computing environments which consists of two main components, namely policy management and specification component and policy evolution component.udIn the policy management and specification component, we first introduce policy management as a service (PMaaS), a cloud based policy management framework that give cloud users a unified control point for specifying authorization policies, regardless of where the data is stored. Then, we present semantic based policy management framework which enables users to specify access control policies using semantic web technologies and helps address heterogeneity issues of cloud computing environments.udWe also model temporal constraints and restrictions in GTRBAC using OWL and show how ontologies can be used to specify temporal constraints.udWe present a proof of concept implementation of the proposed framework and provide some performance evaluation.ududIn the policy evolution component, we propose to use role mining techniques to deal with policy evolution issues and present StateMiner, a heuristic algorithm to find an RBAC state as close as possible to both the deployed RBAC state and the optimal state. We also implement the proposed algorithm and perform some experiments to demonstrate its effectiveness.