Integrated NFV/SDN Architectures: A Systematic Literature Review

Abstract

Network Functions Virtualization (NFV) and Software-De ned Networking (SDN) are new paradigms in the move towards open software and network hardware. While NFV aims to virtualize network functions and deploy them into general purpose hardware, SDN makes networks programmable by separating the control and data planes. NFV and SDN are complementary technologies capable of providing one network solution. SDN can provide connectivity between Virtual Network Functions (VNFs) in a exible and automated way, whereas NFV can use SDN as part of a service function chain. There are many studies designing NFV/SDN architectures in di erent environments. Researchers have been trying to address reliability, performance, and scalability problems using di erent architectural designs. This Systematic Literature Review (SLR) focuses on integrated NFV/SDN architectures, with the following goals: i) to investigate and provide an in-depth review of the state-of-the-art of NFV/SDN architectures, ii) to synthesize their architectural designs, and iii) to identify areas for further improvements. Broadly, this SLR will encourage researchers to advance the current stage of development (i.e., the state-of-the-practice) of integrated NFV/SDN architectures, and shed some light on future research e orts and the challenges faced. CCS Concepts: • General and reference → Surveys and overviews; • Networks → Network design principles; Network services;

Robust network function virtualization

“…We review in this section the works on NFV resource allocations and existing solution approaches and focus on the approximation algorithms for facility location problems and relevant location routing problems. Related topics have been reviewed in survey papers, such as NFV architectures [23], mobile applications [24], and NF deployment/provisioning related resource allocation [25], [26]. Most recent works on NFV MANO focus on NFV instantiation [27], [28], orchestration [29], management [30], and scheduling [31].…”