隨著行動網路演進,在5G應用中提出支援多種不同的網路切片需求,例如eMBB、URLLC與mMTC等不同服務型態的網路切片。也因此在5G定義支援網路功能全虛擬化管理調度平台(NFV MANO),電信營運商可將網路切片中的虛擬化網路功能(VNF)運行於與雲端系統架構一樣的一般商用伺服器(Commodity Server)環境中,網路功能可以以虛擬機或容器方式運行,透過VNF使系統具有快速彈性部署及網路功能故障迅速復原等優點。此外,為進一步自動化優化管理,在5G 3GPP TR 28.861文件中提出自我組織網路(SON)設計,其中self-optimization用於網路切片自我優化,需滿足負載平衡、可靠性優化以及省電等需求。而5G網路隨著支援各種網路切片的服務,將有大量的VNFs運行並進行各項管理與調度,節能的需求變得重要。因此,本論文基於NFV MANO平台,研究調度放置網路切片相關VNFs優化策略,達到SON自動化節能自我優化的需求。
As mobile networks evolve, 5G applications have been proposed to support a variety of different network slicing requirements. For example, network slices of different service types include eMBB, URLLC, mMTC, etc. Therefore, an NFV management and network orchestration (NFV MANO) architecture supporting fully virtualized network functions is defined in 5G. Telecom operators can run the virtualized network function (VNF) in network slicing in the same commodity servers environment as the cloud system architecture. Network functions can be run in the form of virtual machines or containers. Through VNF, the system has the advantages of rapid and flexible deployment and rapid recovery of network function failures. In addition, in order to further automate optimization management, self-organizing network (SON) design is proposed in the 5G 3GPP TR 28.861 document. Among them, self-optimization is used for network slice optimization, which needs to meet load balancing, reliability optimization, and energy-saving. As the 5G network supports various network slicing services, there will be a large number of VNFs running and performing various management and scheduling, and the demand for energy-saving becomes important. Therefore, this study is based on the NFV MANO platform, to study the optimization strategy of VNFs related to scheduling and placement of network slices to achieve SON design meets the needs of automatic self-optimization for energy-saving.