Towards alignment of architectural domains in security policy specifications

Abstract

Large organizations need to align the security architecture across three different domains: access control, network layout and physical infrastructure. Security policy specification formalisms are usually dedicated to only one or two of these domains. Consequently, more than one policy has to be maintained, leading to alignment problems. Approaches from the area of model-driven security enable creating graphical models that span all three domains, but these models do not scale well in real-world scenarios with hundreds of applications and thousands of user roles. In this paper, we demonstrate the feasibility of aligning all three domains in a single enforceable security policy expressed in a Prolog-based formalism by using the Law Governed Interaction (LGI) framework. Our approach alleviates the limitations of policy formalisms that are domain-specific while helping to reach scalability by automatic enforcement provided by LGI.