An Efficient VNF Deployment Algorithm for SFC Scaling-out Based on the Proposed Scaling Management Mechanism

Abstract

Network Function Virtualization (NFV), as an emerging network virtualization technology, replaces hardware function devices by virtualized network functions (VNF). NFV realizes flexible and rapid deployment and maintenance for network functions. On this basis, the service function chain (SFC) satisfies users' demands by connecting various VNFs. To adapt to the dynamic changes in service demand and traffic, VNF scaling guarantees the user experience by increasing the resources allocated to the network when the network capacity is overloaded, also reduces resources when resource utilization is low in the network. Most of the research in this field focuses on the effectiveness of VNF deployment when creating SFC. However, there also remains problems of high transmission delay and low resource utilization of the physical machine (PM) for VNF deployment after scaling-out. Therefore, this paper proposed a scaling management mechanism and optimized VNF deployment algorithm, which can reduce the transmission delay and improve the resource utilization of the computing network on the basis of achieving VNF scaling. Through simulation and measurement, the experimental results show the advantages of our proposed algorithm in reducing transmission delay and improving resource utilization by increasing non-started PMs.