Proactive restoration of optical links based on the classification of events

With the recent explosion of Internet traffic, bandwidth in optical core networks has greatly increased. Because of the huge volume of transported data and services, survivability in these networks is a big issue. Technical solution for securing this traffic takes advantage of reactive methods such as protection and restoration for rerouting the traffic in a case of a failure. Among the different reasons for failures, an important percentage is actually due to optical cables damage or cuts. In this paper, we propose a new method of proactive protection/restoration where the schemes are triggered before the optical fiber is cut. The method is based on the classification of the events occurring on the optical fiber. Events are monitored thanks to real time measurement of the optical fiber parameters. Classification is based on a suitable representation of the monitored parameters and a training set made of risky events allowing early detection of its premises. The advantage of the method is discussed with respect to its ability to reduce loss of data in case of a fiber cut and thus to improve network availability. Markov state model for proactive and reactive restoration of optical links has been developed for the availability analysis.