Penguin Search Aware Proactive Application Placement

Abstract

With the huge proliferation of IoT devices, new challenges have been raised. These IoT devices generate a huge amount of data, instantly. In addition, they are time sensitive, geographically distributed, require high bandwidth, and location awareness. In order to cope with these challenges, recent studies have allowed exploring a new paradigm so-called fog computing. This latter extends Cloud computing at the edge of the network. Fog computing is an intermediate layer that facilitates the deployment of IoT applications by leveraging new characteristics such as support of mobility, location-awareness, and lower latency. However, its limited resources arise the problem of resource provisioning which has an impact on the application placement decisions. In this paper, we focus on the mobile application placement problem in hybrid Cloud-Fog environment. We have considered both delay-sensitive and delay-tolerant applications. Hence, we propose an exact solution as well as a new approach based on penguin search metaheuristic named PsAAP to fulfill the dynamic demands as well as the application’s QoS requirements. To evaluate the proposed approach, we introduce a mobile scenario including three different types of applications. Moreover, we compare the suggested policy with the exact solution, baseline algorithms, heuristic, and metaheuristic methods. Experiments have been conducted using CPLEX and IfogSim-simulator. The final results show the effectiveness of the proposed approach.