Multi-objective approach of energy efficient workflow scheduling in cloud environments

摘要

Scheduling the tasks of aworkflow to the cloud resources is awell-knownN-P hard problem. Thestakeholders involved in a cloud environment have different interests in scheduling problem. Inaddition to the traditional objectives likemakespan, budget, and deadline, optimized in workflowscheduling, considering the green aspect of cloud, (ie, energy consumption) increase the problemcomplexity. Moreover, the interests of a cloud's stakeholders are conflicting, and satisfying allthese interests simultaneously is a big problem. In this paper, we proposed a new Multi-ObjectiveGenetic Algorithm(MOGA) for workflow scheduling in a cloud environment. MOGA consideredthe conflicting interest of the cloud stakeholders for optimization and provided a solution, whichnot onlyminimizes themakespan under the budget and deadline constraints but also provided anenergy efficient solution using the dynamic voltage frequency scaling.We provided a gap searchalgorithm in this paper, which is used to optimize the resource utilization of the cloud's resources.We compared our results with genetic algorithms considering the budget, deadline, and energyefficiency individually.We also compared the performance of MOGA with Multi-objective ParticleSwarm Optimization (MOPSO)with the same objectives as those ofMOGA. To the best of ourknowledge, there is no solution presented in the literature that considers the diverse objectivesconsidered in this work. The results show that our proposed algorithm MOGA has significantlyimproved not only in terms of budget, deadline, and energy but also improved the utilization ofcloud's resources as compared to the competitive algorithms of this work.