Efficient Upgradation Strategies For ROADMs In WDM Networks

Reconfigurable optical add/drop multiplexer(ROADM)is an important component of wavelength division multiplexing(WDM)optical network.At present,a large number of C ROADM nodes have been deployed in the WDM networks.However,directional constraint of C ROADMs affects WDM network performance in terms of lightpath provisioning.Benefiting from the development of optical device,more advanced Colorless and Directionless or Contentionless(CD/CDC)ROADM architectures have been proposed.To meet the rapid growth of network traffic demands,network carriers start to upgrade optical network nodes with CD/CDC ROADMs for more flexible lightpath reconfiguration capabilities.This therefore leads to an important problem of ROADM node upgradation.In this paper,we explore this problem under both the static traffic demand and dynamic traffic demand.Under static traffic demand,we firstly defines this ROADM node upgradation problem in a WDM network.Then,with the objective of maximizing the satisfied demand units,we developed an integer linear programming(ILP)model.Considering high computational complexity of the ILP model,this paper also proposed three node upgradation strategies,including random,demand distribution(DD),and route density(RD)based strategies.Simulation results show that the results of proposed RD strategy is close to the ILP model and significantly better than the other two strategies.We also evaluated the impact of the maximum number of available wavelengths on each link and the maximum number of available ports on each ROADM,respectively.We observed that the impact of CDC ROADMs on satisfied demand units significantly decrease with an increasing maximum number of available wavelengths on each link.The impact of CDC ROADMs on satisfied demand units increases with an increasing maximum number of available ports on each ROADM.Under dynamic traffic demand,this paper also defines the ROADM node upgradation problem.We consider three different node upgradation strategies,including random,route density(RD),and blocking reason(BR)based strategies.The simulation results show that BR strategy performs best,which can reduce the lightpath blocking probability by up to 39.5%in comparison with the RD strategy.We also evaluated how the lightpath blocking performance changes with the increase of the maximum number of available wavelengths on each link and the maximum number of available ports on each ROADM after ROADM node upgradation,respectively.Similar to results under static traffic demand.The impact of CDC ROADM nodes decrease with the increase of the maximum number of available wavelengths on each link and increase with the increase of the maximum number of available ports on each ROADM.