Resource-Efficient Virtual Network Function Placement in Operator Networks

Abstract

Network Function Virtualization (NFV) is an emerging network resource utilization approach which decouples network functions from proprietary hardware. To accommodate Service Function Chain (SFC) requests, service providers offer Virtual Network Function (VNF) instances in operator networks. However, how to efficiently place VNFs at various network locations while jointly optimizing computing and communication resource is still an open problem. To this end, we study the resource-efficient VNF placement problem in operator networks. We firstly formulate this problem as an Integer Linear Programming (ILP) model. Then we design an efficient heuristic algorithm named Resource- efficient Virtual Network Function Placement (RVNFP) based on Hidden Markov Model (HMM). Extensive simulation results show that compared with the previous VNF placement algorithms, RVNFP saves up to 12.51% network cost, and achieves a good tradeoff between computing resource cost and communication resource cost.