Expect the unexpected: Towards a middleware for policy adaptation in IoT platforms

Abstract

Internet of Things platforms are used increasingly to connect devices, applications, and users. Internal and external events, such as updates and calamities, may alter the policies that are actively enforced to meet quality and functional requirements. While many of the events resulting in policy change can be anticipated, exceptional situations such as a terrorist attack urge the need for rapid dynamic installation and activation of response policies. In this paper, we classify events based on expectedness of time of occurrence and of response measures. Furthermore, we examine the requirements of a policy system that enforces qualities and behaviour in an IoT platform based on adaptive policies for expected and unexpected events. We evaluate the impact of these expectedness levels on a hierarchical conceptual policy system model.