Our paper titled “FEAST: A Framework for Evaluating Implementation Architectures of Self-adaptive IoT Systems” was accepted in the 10th technical track on software architecture (SA-TTA 2022) as a part of The 37th ACM/SIGAPP Symposium On Applied Computing. This paper came out as a result of the project done by Federico Di Menna.

Authors: Federico Di Menna, Henry Muccini and Karthik Vaidhyanathan

Abstract

Self-adaptation is considered as one of the effective means to enable IoT systems to overcome run-time uncertainties and thereby provide guarantees on the Quality of Service (QoS). Traditionally this is accomplished by extending the IoT system (managed system) with a managing system responsible for handling the adaptation. Many different approaches have been proposed in the literature to perform such adaptations but, they mainly focus on improving the adaptation logic inside the managing system to provide better guarantees on the QoS of the managed system. However, they do not focus on the implementation architecture of the managing system itself. Further, the choice of implementation can impact the QoS offered by the IoT system and that of the managing system. To this end, in this work, we propose a framework that allows the architect to systematically design, analyze and evaluate different architectural strategies for implementing the managing system of self-adaptive IoT systems. The evaluation results of our approach on the case study of a real-world IoT system demonstrate promising results in allowing architects to evaluate multiple architectural strategies for implementing the managing system.