Virtual network function intelligent migration method

Abstract

The invention relates to a virtual network function intelligent migration method, and belongs to the technical field of mobile communication. In the virtual network function intelligent migration method, on each discrete time slot, on the premise that the highest average time delay constraint, the node cache resource consumption limitation and the link bandwidth capacity limitation of each slice are ensured, according to the queue state information, the node state information and the link state information of each Virtualized Network Function (VNF) instance, an optimal VNF migration strategy is formulated for slices, and an allocation strategy of node CPU resources in the network is dynamically adjusted by aiming at minimizing the average operation energy consumption of the general server. According to the virtual network function intelligent migration method, CPU resources can be fully utilized, and the requirement for average end-to-end time delay of slices can be met; and the sharing of the VNF instances is realized, and the stability of the system is effectively maintained while the average energy consumption of the system is saved.

Description

technical field [0001] The invention belongs to the technical field of mobile communication, and relates to a virtual network function (Virtualized Network Function, VNF) intelligent migration method. Background technique [0002] 5G network slicing and NFV technology bring great advantages to the centralized scheduling of virtual resources and the flexible orchestration of VNFs. Under the NFV architecture, traditional network elements are abstracted into independent functional modules in the form of software, and deployed on a common server platform; at the same time, multi-dimensional network resources are integrated into a unified resource pool, and different slices are allocated according to users. The resources in the resource pool are dynamically scheduled and allocated on demand. This mode makes it possible to manage the mobility of SFC in the infrastructure. As the network load and the status of links and nodes in the infrastructure change, the deployment mode of th...

AI Technical Summary

Problems solved by technology

[0004] In the current literature on VNF migration, most of the work stays on the problem of instantaneous index optimization in a fixed network environment, that is, single-slot optimization. For dynamic slicing business scenarios with a certain life cycle, research uses online There is not much work on resource management techniques to optimize the average performance metrics of VNF migration over long time scales

In addition, in the shared state of VNF instances, that is, when a single VNF instance can be scheduled by multiple slice services at the same time, if the performance of a certain slice cannot meet the business requirements, the current research method cannot formulate an effective strategy to realize the migration of the service function chain VNF

Images