| Network slice has become one of the core technologies of 5G network,which provides the matching network function for different service scenario through virtualization technology.However,if the tenant resources in network slice are statically allocated,due to the dynamic nature of the network status and service traffic,the resource allocation among the tenants in the slice is uneven,resulting in insufficient resource utilization.Therefore,it needs designing a flexible and dynamic method for resource allocation in network slice to maximize revenue of network resources,which is the primary problem solved in this paper.In this paper,based on the analysis of existing problems of long service recovery time when improving average resource utilization on resource dynamic allocation scheme based on SDN,it proposes a method of dynamic resource adjustment of the tenant network in network slice(DRANS)based on SDN.Firstly,it establishes the resource dynamic adjustment trigger mechanism,the resource utilization upper and lower thresholds and timeout time are set to control the frequency granularity of resource adjustment.Secondly,it establishes an on-demand resource allocation mechanism.Building a benefit model to calculate the resource allocation value for each tenant,realize that minimize adjustment overhead and maximize the resource utilization.The purpose of the mechanism is to maximize the revenue of network slice resource.Thirdly,introducing updated number of forwarding rule as a cost,it builds the multi-commodity flow model to search the optimal link to map.Thus,reducing the service recovery time.Finally,the resource dynamic adjustment module is developed in Ryu controller.Experiments were carried out for both high-load and light-load traffic flow scenarios.Through the experimental data,it shows that the method can increase the average resource utilization of each tenant resource by 5%and can save a flow table update time when services recovery. |
PDF Full Text Request