Distribution of virtual devices on the fog for delay and araffic aeduction

Abstract

By emerging Internet of Things (IoT) applications, in near future billion of devices will be connected to the internet. Because of variations in IoT standards, protocols, and devices and also the limited resources of devices, many studies have suggested exploiting virtual devices that perform computation and communication tasks on behalf of physical devices. However, locating virtual devices on far clouds will increases communication delay of devices and overall internet traffic. For this reason, in many studies, the fog computing is used for offloading devices' tasks on resources that are located near the physical device. In this paper, we propose a way to assign virtual devices to fog resources to minimize total delay of communication between physical devices with their corresponding virtual devices. Because computational complexity of the proposed virtual device assignment problem like a generalized assignment problem is NP-Hard, in this paper we propose a genetic algorithm for obtaining near-optimal solutions. First, for solving the problem we use the clustering algorithm, then we improve the results by using proposed genetic algorithm. Simulation results shows that the genetic algorithm improves the results 8.2 percent.