Multilayer Network Analytics With SDN-Based Monitoring Framework

Abstract

This is the author accepted manuscript (AAM). The final published version (version of record) is available online via IEEE at https://www.osapublishing.org/jocn/abstract.cfm?uri=jocn-9-2-A271. Please refer to any applicable terms of use of the publisher. University of Bristol -Explore Bristol Research General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Full terms of use are available: http://www.bristol.ac.uk/pure/about/ebr-terms > REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 1  Abstract-Optical transport networks with expanding variety and volume of network data services challenge network provider's ability to provide high-quality service assurance and network managements. Software-defined networks (SDN) decouple the data plane and control plane and enable network programmability in optical networks with a centralized network controller. The optical network would become more dynamic in network architecture and require frequent network reconfigurations. The dynamic optical networks require all kinds of visibility into application data types, traffic flows and end-to-end connections. Thus, we propose an SDN-based monitoring framework that expands network analytics to a converged packet and optical network. With a designated monitoring hub, monitoring information from multiple layers are collected and processed in a centralized server. Several monitoring technologies are provided as network services with the architecture-on-demand optical node architecture. The developed network applications on top of SDN controller process all the collected monitoring information and enables multilayer network analytics based the SDN-based monitoring framework. We demonstrate the proposed multi-layer network analytics successfully in several use cases. In the demonstrations, the multi-layer network analytics enables QoS recovery to avoid network disruption, optical power equalization at any combining device and network debugging and restoration in optical networks. The developed multilayer network analytics provides powerful tools for network re-planning and optimization dynamically.

Stateful firewall‐enabled software‐defined network with distributed controllers: A network performance study

“…Dedicated monitoring devices collect all the monitored information from multiple layer of the SDN; the information is processed to maintain network stability. Feature like power management, QOS, network preplanning, network optimization, network performance, network debugging, and restoration can be enhanced by analyzing and diagnosing with multilayer network monitoring framework . Routing performance on distributed control plane is enhanced by using divide and conquer technique.…”

NFVMon: Enabling Multioperator Flow Monitoring in 5G Mobile Edge Computing

“…Packet classification and metric calculation are done directly on the switch and metrics are then sent to the traffic analyzer application running in the SDN controller. Yan et al [23] propose multilayer network analytics architecture based on SDN-based to carry out monitoring of metrics related to the physical layer. These metrics are sent to the SND controller in order to control link impairments, link performance, and latency.…”

Field trial of Machine-Learning-assisted and SDN-based Optical Network Planning with Network-Scale Monitoring Database

“…Network Planning with Network-Scale Monitoring Database Shuangyi Yan (1) , Faisal Nadeem Khan (2) , Alex Mavromatis (1) , Dimitrios Gkounis (1) , Qirui Fan (2) , Foteini Ntavou (1) , Konstantinos Nikolovgenis (1) , Fanchao Meng (1) , Emilio Hugues Salas (1) , Changjian Guo (3) , Chao Lu (3) , Alan Pak Tao Lau (2) , Reza Nejabati (1) and Dimitra Simeonidou (1) (1) High Performance Networks Group, University of Bristol, shuangyi.yan@bristol.ac.uk (2) Photonics Research Center, Department of Electrical Engineering, The Hong Kong Polytechnic University (3) Photonics Research Center, Department of Electronic and Information Engineering, The Hong Kong…”

“…In addition, the spectrum information at some nodes are also be stored. On top of the database, multiple data analytics applications can run in real-time 3 . Machine-learning based network abstraction OSNR have been used as an indicator for QoT.…”