Optimal approach for electronic grooming, routing and spectrum allocation in elastic optical networks

Abstract

This work proposes a new Integer Linear Programming (ILP) formulation for traffic grooming, routing and spectrum allocation (GRSA) to be used in translucent elastic networks. In order to compare our method with classic routing and spectrum allocation (RSA), it was presented an ILP model for RSA problem, allowing the evaluation of the benefits of traffic grooming in translucent elastic networks. Numerical experiments were accomplished for 6-, 10-, and 14-node topologies. In addition, it was made 4 simulations for each topology: for each model (with and without traffic grooming) two objective functions were tested (minimizing global frequency slots utilization and minimizing the congested physical link). Results exposed the benefits of traffic grooming in elastic networks, achieving until 60% of frequency slots reduction, when compared with a network design without traffic grooming in a 14-node network. Moreover, it was presented the relation between network configuration characteristics and the objective function used.