On the Performance of Fog-Cloud Computing for Real-time Surveillance Applications

With the remarkable growth of IoT networks, there is a need for viable infrastructures that enable the successful deployment of IoT-based applications. Recently, cloud computing Postured itself as a dynamic promising structure to transfer and process the large amount data generated by the IoT nodes through the Internet. However, the applications are of stringent requirements, like video applications, the cloud fails to meet these requirements due to the latency problems experienced during heavy Internet traffic. Therefore, a more promising infrastructure is fog computing, which brings the cloud closer to the user through intermediary devices known as fog nodes. Therefore, the purpose of this paper is to study the viability of fog computing as a medium for IoT-based real-time surveillance applications. Specifically, we investigate and compare the performance of a classical cloud computing paradigm and a hybrid fog-cloud architecture. Our simulations shows that the hybrid model outperforms the classical model in terms of the number of requests meeting their deadlines and in terms of the forward trip time.