Research on Intelligent Management Scheme of Network Slicing

Abstract

Network slice technology can provide end-to-end virtual private networks which are logically isolated for differentiated services on the basis of the same set of network infrastructure equipment. The differentiation of business requirements, the dynamic business requirements, and the complexity of network infrastructure have brought great challenges to the on-demand customization of network slicing, and the release, redistribution of network resources. It is proposed to use the data of each service as a data set to train the sequence prediction model and the Service Level Agreement(SLA) parameter-Virtual Network Function(VNF) mapping model by distributed Artificial Intelligence(AI) to predict the traffic change of each subnet slice and the management of the slice, and then manage each subnet slice in a differentiated and intelligent pattern.