An iterative greedy algorithm for scheduled traffic grooming in wdm optical networks

Routing and wavelength allocation (RWA) in translucent WDM optical networks aims at optimizing network's resources utilization. Traffic grooming, traffic rerouting, wavelength conversion and regenerator placement are some of the key operations enabling such an optimization. Two types of traffic grooming are considered in WDM networks, either optical or electrical. In this paper, we propose an innovative greedy iterative algorithm enabling to aggregate in the electrical domain scheduled traffic demands in order to favors a better bandwidth utilization of the optical channels to be set-up in the network. For a same set of traffic scenarios, we compare the cost of the resources required in the network, either by means of our algorithm or by means of two other greedy iterative algorithms recently proposed in the literature. The optical fiber infrastructure being installed, we consider the network cost as proportional to the global number of required optical ports and electrical ports at the various switching nodes.