A VNF Placement Method Considering Load and Hop count in NFV Environment

Abstract

According as the age of the fourth industrial revolution has arrived, the virtualization technology is in the spotlight, and is greatly influencing changes in IT environment. Particularly, NFV (Network Function Virtualization), which applies the technologies of virtualization to the area of networks, separates software functions from the hardware-dependent network devices, components of the existing network, and provides services through the virtualization of general purpose server equipment-based infrastructure, and is drawing attention as the key technology in the recent 5G environment. Functions that are virtualized and operate in the above NFV environment are called VNFs (Virtualized Network Functions), and research is being carried out about how to deploy VNFs efficiently, which is the biggest issue in the NFV environment. The optimized VNF placement refers to the optimized mapping of VNF requirements (e.g. the number of CPU cores, memory, network bandwidth, etc.) to the physical node that can provide relevant resources. This paper suggests a method that can solve, through the threshold value, the load caused by the indiscreet generation of virtualized instances arising from using only simple shortest path algorithms as well as deploy efficiently VNFs in the NFV environment by using shortest path algorithms.