Intent-based zero-touch service chaining layer for software-defined edge cloud networks

Edge Computing, along with Software Defined Networking and Network Function Virtualization, are causing network infrastructures to become as distributed clouds extended to the edge with services provided as dynamically established sequences of virtualized functions (i.e., dynamic service chains) the...

Full description

Saved in:
Bibliographic Details
Published inComputer networks (Amsterdam, Netherlands : 1999) Vol. 212; p. 109034
Main Authors Martini, B., Gharbaoui, M., Castoldi, P.
Format Journal Article
LanguageEnglish
Published Amsterdam Elsevier B.V 20.07.2022
Elsevier Sequoia S.A
Subjects
Adaptation
Automation
Chaining
Chains
Cloud computing
Delivery services
Edge computing
Edge networking
IBN
NFV
Operators (mathematics)
SDN
Service chaining
Software
Software-defined networking
Service chaining
IBN
SDN
Edge networking
NFV
Online AccessGet full text

Cover

More Information
Summary:Edge Computing, along with Software Defined Networking and Network Function Virtualization, are causing network infrastructures to become as distributed clouds extended to the edge with services provided as dynamically established sequences of virtualized functions (i.e., dynamic service chains) thereby elastically addressing different processing requirements of application data flows. However, service operators and application developers are not inclined to deal with descriptive configuration directives to establish and operate services, especially in case of service chains. Intent-based Networking is emerging as a novel approach that simplifies network management and automates the implementation of network operations required by applications. This paper presents an intent-based zero-touch service chaining layer that provides the programmable provision of service chain paths in edge cloud networks. In addition to the dynamic and elastic deployment of data delivery services, the intent-based layer offers an automated adaptation of the service chains paths according to the application’s goals expressed in the intent to recover from sudden congestion events in the SDN network. Experiments have been carried out in an emulated network environment to show the feasibility of the approach and to evaluate the performance of the intent layer in terms of network resource usage and adaptation overhead.
ISSN:1389-1286
1872-7069
DOI:10.1016/j.comnet.2022.109034