A restoration framework for partial failures in wireless networks - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2015

A restoration framework for partial failures in wireless networks

Résumé

The purpose of the paper is two-fold. First, we introduce a general framework for extending restoration strategies developed for total link failures in a way enabling handling partial link failures as well. In particular, our framework combines a known rerouting method for total link failures (the so called Restricted Restoration - RR) and a recently developed restoration method for partial failures (called Flow Thinning - FT). In this way we achieve two goals: we define a unified approach to deal with both kinds of failures, and improve efficiency of FT. We study in detail such an extended version of RR (that can also be regarded as an extension of the MPLS Fast Reroute method) capable of managing links with variable capacities, (variable link capacities can be regarded as partial link failures). A general mathematical model for the framework, and a specific model for RR are presented together with numerical results. The second purpose is related to completeness of partial failures scenarios. In practice, partial failure scenarios correspond to partial losses in transmission capacity, occurring for example in wireless networks using microwave radio or free space optical transmission technology. As such failures are mainly caused by weather conditions, there arises a problem of how to translate weather conditions to link failures (this is necessary to ensure network robustness to weather conditions). We present a study on the way that different weather factors impact the transmission link capacity and how these effects can be measured.
Fichier non déposé

Dates et versions

hal-01298056 , version 1 (05-04-2016)

Identifiants

Citer

Yoann Fouquet, Dritan Nace, Michal Pioro, Michael Poss. A restoration framework for partial failures in wireless networks. RNDM: Reliable Networks Design and Modeling, Oct 2015, Munich, Germany. pp.149-155, ⟨10.1109/RNDM.2015.7325222⟩. ⟨hal-01298056⟩
196 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More