Loading…

A meta-heuristic approach for non-bifurcated dedicated protection in WDM optical networks

Survivable routing serves as an important issue in core optical mesh networks. Operators are mainly interested in the protection of single link failures and a set of multiple link failures with high failure probability. In order to consider failure dependencies among links, Shared Risk Link Groups (...

Full description

Saved in:
Bibliographic Details
Main Authors: Soproni, P., Babarczi, P., Tapolcai, J., Cinkler, T., Pin-Han Ho
Format: Conference Proceeding
Language:English
Subjects:
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Survivable routing serves as an important issue in core optical mesh networks. Operators are mainly interested in the protection of single link failures and a set of multiple link failures with high failure probability. In order to consider failure dependencies among links, Shared Risk Link Groups (SRLG) are defined in the network. Against a given set of SRLG failures the connection needs to be resilient to fulfill the required quality of service (QoS) level. Currently, 1+1 dedicated path protection is reported to be the most widely deployed network resilience mechanism because it offers instantaneous recovery from network failures. However, 1+1 protection in most cases unable to provide SRLG-disjoint paths, thus, it is not resilient against all SRLG failures under consideration. In this paper, a novel Bacterial Evolutionary Algorithm for the Integer (or non-bifurcated) Generalized Dedicated Protection (IGDP) is introduced to protect each SRLG in the operator's list. The proposed evolutionary algorithm aims to minimize the amount of reserved spare bandwidth. As a result, the solution is resilient against all failures in the SRLG list while the simplicity of 1+1 protection in signaling and switching is maintained. Extensive simulation is conducted to compare our evolutionary approach with previously reported optimal and heuristic solutions for the IGDP problem, by which its merits are further demonstrated.
DOI:10.1109/DRCN.2011.6076892