Load Aware Management of Cloudlets for a Wireless Area Metropolitan Network

Abstract

Wireless metropolitan area network (WMAN) emerges to be popular and powerful to provide abundant access services for mobile devices in the construction of smart city. To improve the quality of experience (QoE) for mobile users in WMAN, cloudlet is treated as a promising paradigm which provides potential extension of the limited resources to the mobile devices. However, it is still a challenge to realize the dynamic load-balance distribution of the performed mobile applications for cloudlet management in WMAN. In view of this challenge, a Load-aware Resource Scheduling method, named LRS, is designed to achieve the goal of load balancing. Technically, we present the load balance analysis for the cloudlets in WMAN first. Then a corresponding virtual machine (VM) scheduling method is developed to adjust the resource provision dynamically by leveraging the VM live migration technique. Finally, extensive experimental evaluations are conducted to verify the effectiveness and efficiency of our proposed method.