Visible to the public Guaranteeing QoS requirements in long-haul RINA networks

TitleGuaranteeing QoS requirements in long-haul RINA networks
Publication TypeConference Paper
Year of Publication2017
AuthorsLeon, S., Perelló, J., Careglio, D., Tarzan, M.
Conference Name2017 19th International Conference on Transparent Optical Networks (ICTON)
Date Published July 2017
Keywordsbuilt-in QoS support, clean slate, Collaboration, Computer architecture, connection-oriented solutions, data transport networks, Degradation, fastest-path solutions, flow QoS parameters, heterogeneous Quality of Service requirements, Human Behavior, human factor, human factors, Internet, long-haul, long-haul RINA networks, maximum allowed latency, Measurement, Metrics, network providers, network requirements, packet losses, path selection, policy governance, Policy-Governed Secure Collaboration, programmable networking model, Protocols, pubcrawl, QoS, QoS requirements, quality of service, Recursive InterNetwork Architecture, recursive networking model, resilience, Resiliency, RINA, RINA architecture, shortest-path solutions, static layered structure, TCP-IP protocol stack limitations, TCP-IP-based Internet, TCPIP, telecommunication network topology, traffic flows, transport protocols
Abstract

In the last years, networking scenarios have been evolving, hand-in-hand with new and varied applications with heterogeneous Quality of Service (QoS) requirements. These requirements must be efficiently and effectively delivered. Given its static layered structure and almost complete lack of built-in QoS support, the current TCP/IP-based Internet hinders such an evolution. In contrast, the clean-slate Recursive InterNetwork Architecture (RINA) proposes a new recursive and programmable networking model capable of evolving with the network requirements, solving in this way most, if not all, TCP/IP protocol stack limitations. Network providers can better deliver communication services across their networks by taking advantage of the RINA architecture and its support for QoS. This support allows providing complete information of the QoS needs of the supported traffic flows, and thus, fulfilment of these needs becomes possible. In this work, we focus on the importance of path selection to better ensure QoS guarantees in long-haul RINA networks. We propose and evaluate a programmable strategy for path selection based on flow QoS parameters, such as the maximum allowed latency and packet losses, comparing its performance against simple shortest-path, fastest-path and connection-oriented solutions.

URLhttp://ieeexplore.ieee.org/document/8025071/
DOI10.1109/ICTON.2017.8025071
Citation Keyleon_guaranteeing_2017