An energy-optimized embedded load balancing using DVFS computing in cloud data centers

Abstract

Task scheduling is a significant challenge in the cloud environment as it affects the network’s performance regarding the workload of the cloud machines. It also directly impacts the consumed energy, therefore the profit of the cloud provider. This paper proposed an algorithm that prioritizes the tasks regarding their execution deadline. We also categorize the physical machines considering their configuration status. Henceforth, the proposed method assigns the jobs to the physical machines with the same priority class close to the user. Furthermore, we reduce the consumed energy of the machines processing the low-priority tasks using the DVFS method. The proposed method migrates the jobs to maintain the workload balance, or if the machines’ class changed according to their scores. We have evaluated and validated the proposed method in the CloudSim library. The simulation results demonstrate that the proposed method optimized energy consumption by 12% and power consumption by 20%.