Double-link failure localization based on the monitoring trail concept in WDM networks - IMT - Institut Mines-Télécom Accéder directement au contenu
Communication Dans Un Congrès Année : 2012

Double-link failure localization based on the monitoring trail concept in WDM networks

Résumé

As traffic demands are continuously increasing, core networks must be able to react almost instantaneously to any single, even multiple, failure(s) in order to prevent huge data losses. Even though the probability of simultaneous multiple failures is very small, the time needed to repair a single failure ranges from several hours (for landline fibers) to several weeks (for submarine fibers). During this period, the network is vulnerable to any new failure that may occur in the network. Therefore, designing survivable WDM networks against multiple failures requires efficient and accurate detection and localization mechanisms. In this paper, we extend the monitoring trail concept in order to detect and unambiguously localize any single- and double- link failure in the network. Compared to our previously proposed MEMOTA algorithm, the improved MEMOTA++ algorithm benefits from an improved trail reconfiguration algorithm in order to reduce its execution time. Numerical simulations have been carried out using the Deutsche Telekom and the Geant-2 European networks. We show that a monitoring solution able to localize any double-link failure in the Deutsche Telekom network is 122% more expensive than the one able to localize any single-link failure, but remains 10.9% more economical than the traditional link-based monitoring solution.
Fichier non déposé

Dates et versions

hal-01326281 , version 1 (03-06-2016)

Identifiants

Citer

E. A. Doumith, Sawsan Al Zahr. Double-link failure localization based on the monitoring trail concept in WDM networks. IEEE International Symposium on Broadband Networks and Fast Internet, May 2012, Baabda, Lebanon. pp.21-28, ⟨10.1109/RELABIRA.2012.6235087⟩. ⟨hal-01326281⟩
44 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More