Modélisation et simulation des services IoT dans un environnement de Fog computing

Abstract

The IoT is a revolutionary technology offering significant benefits in operational efficiency, decision-making and customer satisfaction. However, the effective operation of the IoT requires a high-performance infrastructure, such as that of cloud computing, to manage data and services, as well as application requirements. Cloud computing, however, reveals certain limits related in particular to latency, bandwidth and operating costs. Fog computing is therefore proposed as a solution to this type of problem by offering a distributed computing infrastructure to process data and services close to users and connected objects, thus offering an effective alternative to Cloud computing. In this thesis, we study a model of an IoT system in a Fog computing environment by performing a simulation of IoT services placement strategy on a set of Fog nodes. This strategy is implemented by proposing a "Best Fog" placement algorithm that uses an approach based on several criteria, such as memory and computational capacity, to select the most suitable fog node for a specific task. The simulation results obtained are considered interesting since they reveal a better use of available resources and a more efficient assignment of tasks. This evaluation is also validated by a comparative analysis between the proposed algorithm (“Best Fog”) and another FCFS algorithm.