Combined Latency-Aware and Resource-Effective Virtual Network Function Placement

Abstract

With the emergence of network function virtualization (NFV) and Software-Defined Networking (SDN) technologies, virtual network functions (VNF) can be interconnected to deliver different mobile services in 5G communication networks. Our primary purpose is to find the optimal VNF placement reducing resource consumption while providing specific latency and throughput for slicing services. We consider VNFs as M/M/1/C queues chained together to serve customer requests. The problem relies on finding an efficient orchestration and placement of VNFs. In this context, we propose a decision algorithm based two-procedures, the first one, named orchestration phase, aims to manage the reuse of VNFs having the same required functions through a dynamic logit method, and the second one is related to the new placement of VNFs. In this paper, we focus on a simple scenario of delivering video streaming service traversing a standard chain of four VNFs. Simulation results prove the performance of our proposed algorithm in terms of End-to-End delay and dropping probability compared to greedy and affinity algorithms.