Leveraging Machine Learning in SDN/NFV-Based Networks
摘要截稿:
全文截稿: 2019-06-01
影响因子: 11.42
期刊难度:
CCF分类: A类
中科院JCR分区:
• 大类 : 计算机科学 - 1区
• 小类 : 工程:电子与电气 - 1区
• 小类 : 电信学 - 1区
Overview
A key trend of current network evolution is towards network softwarization. The softwarization/virtualization technology aims to enable a network to be programmable in a way that makes the network more flexible, scalable, and reliable, and in turn leads to agile service deployment, low capital and operational expenses, and having self-x properties. Thus far, two widely adopted solutions have been Software Defined Networks (SDN) and Network Function Virtualization (NFV). Both SDN and NFV have become key enabling technologies for 5G networks and can be widely applicable to a range of important domains, including cloud datacenters, IoT, mobile edge computing (MEC), smart grid, cognition-based networks.
Although SDN and NFV facilitate the flexibility and scalability of network services and make the deployment of network services faster and cheaper, such software-based solution also introduces new problems, including throughput performance degradation and unstable jitter. More specifically, in SDN/NFV-based networks, traffic engineering, resource management, and network security are among the challenges that telecommunication service providers need to overcome in order to provide better services to users and improve their revenue. Meanwhile, machine learning (ML) has seen great success in solving problems from various domains. It is believed that ML also has high potential in addressing the aforementioned challenges in SDN/NFV-based networks, especially in the elastic deployment of virtual network functions (VNFs), dynamic service provisioning, adaptive traffic control, and the security issues, as ML is a technology that can effectively extract the knowledge from data, and then accurately predict future resource requirements of each virtualized software-based appliance and future service demands of each user. Though researchers and practitioners have started their research on exploring various ML techniques for leveraging the performance of these virtualized networks, a great deal of challenges are yet to be addressed.
The aim of this special issue is thus to provide a forum for recent research results on the topics relevant to the technological challenges of leveraging ML technology in SDN/NFV-based networks. We solicit high-quality original research works on various aspects of leveraging the performance of SDN/NFV-based networks with ML. Topics of interest include, but are not limited to:
- Resource management in SDN/NFV-based networks with ML technology
- Traffic engineering in SDN/NFV-based networks with ML technology
- Elastic VNF placement and orchestration with ML technology
- Energy efficiency in SDN/NFV-based networks with ML technology
- VNF performance degradation and correction with ML technology
- Security, Privacy, and Trust issues in SDN/NFV-based networks with ML technology
- Novel and innovative machine learning methods for SDN/NFV-based autonomic networks