Abstract
As a new way to design, deploy and manage networkservices, network functions virtualization (NFV) decouplesthe network functions, from one or more physical networkinfrastructures and black boxes so they can run in software.It therefore comes as no surprise that NFV originated fromservice providers, who were looking to improve the deploymentof new network services to support their revenue and growthobjectives. Within the NFV ecosystem, high availability and lowlatency are one of the key QoS benefits that service providerscan expect from the 5G Cloud and the NFV networks to makedelay-critical services such as remote surgery a reality. Therefore,network services should be placed, chained and routed throughthe network considering users/tenants stringent quality of service(QoS) and service-level agreement (SLA) requirements. To thisend, routing and placement optimization plays a major rolein improving network performance and the overall networkcost. In this paper, we study the problem of virtual networkfunctions (VNFs) placement and routing across the physical hoststo minimize overall latency defined as the queuing delay withinthe edge clouds and in network links. In that respect, this worktakes a holistic view by considering not only VNFs chaining andplacement problem but also considering the flows routing aspectsince these two problems are inter-related and have a majorimpact on network latency.
| Original language | English |
|---|---|
| Pages (from-to) | 2346 - 2357 |
| Journal | IEEE Journal on Selected Areas in Communications |
| Volume | 36 |
| Issue number | 10 |
| Early online date | 24 Sept 2018 |
| DOIs | |
| Publication status | Published - Oct 2018 |