Run-time generation, transformation, and verification of access control models for self-protection

Abstract

Self-adaptive access control, in which self-* properties are applied to protecting systems, is a promising solution for the handling of malicious user behaviour in complex infrastructures. A major challenge in self-adaptive access control is ensuring that chosen adaptations are valid, and produce a satisfiable model of access. The contribution of this paper is the generation, transformation and verification of Role Based Access Control (RBAC) models at run-time, as a means for providing assurances that the adaptations to be deployed are valid. The goal is to protect the system against insider threats by adapting at run-time the access control policies associated with system resources, and access rights assigned to users. Depending on the type of attack, and based on the models from the target system and its environment, the adapted access control models need to be evaluated against the RBAC metamodel, and the adaptation constraints related to the application. The feasibility of the proposed approach has been demonstrated in the context of a fully working prototype using malicious scenarios inspired by a well documented case of insider attack.