Poster - Towards Dependable IoT Systems Using Self-Adaptation.

Realizing dependable Internet-of-Things (IoT) comes with numerous challenges. For large-scale IoT deployments, one particular challenge is handling uncertain operating conditions, such as interference in the wireless communication between devices and gateways or changing availability of services. Without proper mitigation of such uncertainties, dependability goals of IoT systems, such as reliability and energy efficiency, may be jeopardized. In our research, we study how self-adaptation can be applied to realize dependable IoT systems regardless of uncertainties. Self-adaptation equips an IoT system with a feedback loop that tracks the uncertainties and adapts the system as needed to maintain its quality goals. Our focus is on large-scale IoT systems with mobile devices. In this paper, we study a typical adaptation problem of IoT where mobile devices need to adapt their power settings while moving in an heterogeneous environment to ensure reliable and energy efficient communication. We evaluate the approach using a simulator of DingNet, an IoT system that is deployed in Leuven, Belgium.