Utilization-aware Virtual Network Function Placement Using NSGA-II Evolutionary Computing

Abstract

Network functions virtualization (NFVs) is a paradigm shift in order to integrate network functions (NFs). Moreover, Virtual Network Functions (VNFs) can easily be transferred from one device to another without the need for a new special hardware installation. On the other hand, VNF placement (VNF-P) is proved to be NP-hard. In this paper we present a multi-objective meta-heuristic solution using the NSGA-II algorithm for the VNF-P problem. The goal of our algorithm is to place VNFs corresponding to different service chains onto physical hosts, in such a way that could increase the physical resource utilization, and reduce the number of used (active) physical hosts. The simulation results using the well-known CloudSim framework show robustness of the proposed method in terms of criteria such as resource utilization, number of used physical hosts and so on.