Abstract
The IP+GMPLS over DWDM model has been considered a trend for the evolution of optical networks. However, a challenge that has been investigated in this model is how to achieve fast rerouting in case of DWDM failure. Artificial Neural Networks (ANNs) can be used to generate proactive intelligent agents, which are able to detect failure trends in optical network links early and to approximate optical link protection mode from 1:n to 1+1. The main goal of this paper is to present an environment called RENATA2 and its process on how to develop ANNs that can give to the intelligent agents a proactive behavior able to predict failure in optical links.
Cite
CITATION STYLE
Gonçalves, C. H. R., Oliveira, M., Andrade, R. M. C., & de Castro, M. F. (2004). Applying artificial neural networks for fault prediction in optical network links. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 3124, pp. 654–659). Springer Verlag. https://doi.org/10.1007/978-3-540-27824-5_88
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
