Dynamic flexgrid optical networks are being developed to operate under highly changing traffic scenarios. Planning such optical networks is fully dependent on the traffic characteristics whereas the constant traffic evolution needs of continuous network planning actions to adapt to new scenarios. Therefore, networks need to be periodically re-planned to consider features such as survivability degree to be reached during its operation. Since traffic changes could lead to the violation of some of these requirements, they can trigger network upgrading process. In this work, we introduce network vulnerability modelling with the aim of providing statistical models to be used as vulnerability predictors in periodical network planning problems.

Peer Reviewed