Cognitive Management for the Internet of Things: A Framework for Enabling Autonomous Applications

Abstract

The handling of the number of objects that will be part of the Internet of Things (IoT) and the various networking technologies used for their interconnection requires suitable architecture and technological foundations. Despite significant work on architectures and test facilities for the IoT, there is still a lack of management functionality to overcome the technological heterogeneity and complexity of the underlying networks and IoT infrastructure, so as to enhance context/situational-awareness, reliability, and energy-efficiency of IoT applications. This article presents a cognitive management framework for the IoT aiming to address these issues, comprising three levels of functionality: virtual objects (VOs), composite VOs (CVOs), and service levels. Cognitive entities at all levels provide the means for self-management (configuration, optimization, and healing) and learning. Three fundamental processes of this framework are presented: dynamic CVO creation, knowledge-based CVO instantiation, and CVO self-healing. A first prototype implementation of this framework and corresponding derived results are presented.