Research On Energy-saving Technology Of Resource Scheduling For Cloud Computing Based On NSGA2

Cloud computing can not only provide on-demand services to Internet users through virtual computerization and distributed computing,but also offer efficient service deployment solutions for service providers.Since cloud computing data centers are widely used,its energy consumption comes into an issue.In order to utilize energy more efficiently,virtual machine(VM)live-migration is adopted to reallocate the resource of the cloud.Previous study on VM scheduling strategies in cloud environment mainly focused on optimizing the utilization of computing resource.However,energy consumption of network resource is increasing and takes a non-negligible part of whole energy consumption.Thus the energy consumption model taking merely computing resource into consider is not proper.In this paper,a VM scheduling strategy based on NSGA2 is proposed,which optimizes the energy consumption of computation and network simultaneously.The specific contributions of this paper are as follows:(1)We combined data center with SDN thus forming an energy consumption model which is used to quantify the energy consumption in the cloud computing data center and forecast after resource load on the basis of previous resource load data.(2)A selection strategy is designed to select the appropriate virtual machine and physical server to migrate according to the utilization of computing and networking resource in datacenter.(3)A virtual machine allocation strategy is designed to optimize the allocation of computing resources and network resources at the same time,so as to enhance the ability of resource aggregation and improve the utilization rate of resources.(4)We realize the online migration of virtual machine between physical servers and analyze the impact of cloud computing data center during virtual machine migration process.(5)We test the performance of the cloud environment energy-saving technology designed by this paper on the simulation platform,and verify the feasibility in the real environment.