Virtual Network Function Placement for Serving Weighted Services in NFV-Enabled Networks

Network function virtualization (NFV) is a novel network architecture yielding many optimization choices to Internet service providers due to its flexibility. Based on this technique, a network service is constructed from a chain of virtual network functions (VNFs). Although many studies deeply inve...

Full description

Saved in:
Bibliographic Details
Published inIEEE systems journal Vol. 17; no. 4; pp. 1 - 12
Main Authors Nguyen, Dung H. P., Lien, Yu-Hui, Liu, Bing-Hong, Chu, Shao-I, Nguyen, Tu N.
Format Journal Article
LanguageEnglish
Published New York IEEE 01.12.2023
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects
Algorithms
Bandwidth
Hardware
Heuristic algorithms
Internet service providers
Micromechanical devices
Network function virtualization
Network function virtualization (NFV)
network resource
NP-hard
Optimization
Placement
Servers
service
Software
Substrates
virtual network function (VNF)
Virtual networks
Online AccessGet full text

Cover

More Information
Summary:Network function virtualization (NFV) is a novel network architecture yielding many optimization choices to Internet service providers due to its flexibility. Based on this technique, a network service is constructed from a chain of virtual network functions (VNFs). Although many studies deeply investigate the NFV and propose approaches to address the problem of deploying services in physical networks under different perspectives, they do not consider the importance of services. This issue becomes more critical once the number of network services ever-increasing while the network resources are limited. Here, each service is assigned a weight representing its importance. From this aspect, maximizing the total weight of services accepted to deploy over the physical network is a critical problem and poses many challenges, particularly for networks with limited resources. Therefore, we call this issue the VNF placement for weighted services (VPWS) problem. We then propose three algorithms termed first-fit server algorithm (FSA), best-fit server algorithm (BSA), and service segment-based algorithm (SSBA), respectively, to address the VPWS problem. Eventually, we perform simulations to evaluate the proposed algorithms and factors that affect the VNF placement. The experiment indicates that the total weight yielded by the SSBA outperforms those of the FSA and BSA.
ISSN:1932-8184
1937-9234
DOI:10.1109/JSYST.2023.3257776