Cooperative Game Theory-Based Approach for Energy-Aware Job Scheduling in Cloud

Document Type


Publication Date



This paper1 addresses the problem of energy-aware job scheduling for underlying cloud nodes using cooperative game theory. The objectives are on resource utilization maximization and the power consumption minimization without violating the job's latest completion time (Makespan). Cloud computing can deliver platform, software, storage and data services through web browsers as a metered service. Due to the skyrocketed electricity cost and a large number of active users, Cloud service providers are highly motivated to adopt a performance guaranteed and cost-effective job scheduler with low power consumption and high job throughput. Therefore, an energy-aware job scheduling algorithm is proposed for a bag of tasks based on the premise of Nash Bargaining Solution (NBS), which can ensure Pareto-optimality. In such a cooperative theoretical gaming, each job seeks to locate a cloud machine that can both guarantee the low energy under certain makespan constraint. Simulation results show that our approach significantly reduces the power consumption by strategically selecting appropriate mapping nodes for prioritized task modules. Our approach consistently achieves lower energy consumption and higher resource utilization than some comparable methods.

This document is currently not available here.