An IoT Reconfiguration Framework Applied Ontology-based Modeling and Bayesian-based Reasoning for Context Management

The rapid evolution in Internet of Things (IoTs), the increasing of the complexity of the sensors, and the dynamic changes and unpredicted events in the IoTs environments pose new challenges to IoTs reconfiguration. In this paper, we propose an IoTs reconfiguration framework, namely ROB-IoT, for the reconfiguration of IoTs applications upon the changes of the complex contexts to improve our previous study. The approach to build the intelligent context management component of ROB-IoT is using ontology for modelling and the Bayesian network for reasoning. A Bayesian network can flexibly handle the uncertain environments of frequent changes in IoTs devices configuration, the proposed framework can adjust to the changing Internet of Things environment, making it more flexible. To validate our proposed framework, experiments have been carried out in the smart campus application by changing the behaviors of the IoTs application. Results show that our proposed framework for improving the timing of re-configured decision making is better for the dynamic reconfiguration of IoTs applications.