Fog‐based semantic model for supporting interoperability in IoT

Abstract

Cloud computing has several favourable characteristics for supporting Internet of things (IoT) applications and services. While provisioning support for heterogeneous devices adapting different semantics in IoT, cloud-based technology is highly used. In semantic model, attaching additional information to raw sensory data is accomplished with the help of ontology. However, the longer distance between the cloud and IoT devices becomes a bottleneck for critical IoT applications. In this study, the authors proposed a fog-based semantic model called semantic-fog to reduce this gap. The proposed model shifts some of the frequently used semantic services of the cloud to the edge of the sensor network. Additionally, it provides an efficient off-loading technique among fog-fog and fog-cloud devices to reduce task execution time and energy consumption of the fog nodes. Furthermore, an effective mapping method for converting raw sensory data into Resource Description Framework format is also presented in this work. A comparative analysis among the relevant cloud-based computing models and the proposed semantic-fog model is performed in terms of service delay, energy consumption, network usage, and the total cost of the system. Simulation results show that semantic-fog outperforms the cloud-based models almost in all the aspects considered.