Autoregressive DragonFly Optimization for Multi-Objective Task Scheduling (ADO-MTS) in Mobile Cloud Computing

Abstract

Mobile Cloud computing is a recognized computing platform and is being considered as a business model due to its potential growth in offering required services to the users. However, it poses a number of challenges, among which the consumption of energy in the data centers is the key issue. Hence, an energy aware task scheduling technique, called Autoregressive Dragonfly Optimization-based Multi-objective Task Scheduling (ADO-MTS), is designed that schedules the tasks to the suitable cloud resources. The scheduling of the tasks is either in the public cloud or Mobile Cloud (MC) such that the utilization of energy is reduced. Accordingly, an optimization algorithm, Autoregressive Dragonfly Optimization (ADO), is developed combining Conditional Autoregressive Value at Risk (CAViaR) with Dragonfly Algorithm (DA). Moreover, a multi-objective model concerning energy consumption, makespan, and resource utilization, is designed for the optimal allocation of resources to the tasks. Three measures, such as resource utilization, makespan, and energy, are used to evaluate the performance of the proposed ADO-MTS technique. The results show that the ADO-MTS technique has achieved the maximum performance by increasing the resource utilization and minimizing the makespan and energy consumption as compared to other existing techniques.