Solving large size instances of the RWA problem using graph partitioning

In this paper, we propose a new heuristic to solve large instances of the Routing and Wavelength Assignment problem (RWA). First, the initial instance is split into several smaller instances that are optimally solved using the ILP formulation of the problem. This decomposition is obtained by partitioning the set of edges of the network and is such that the different instances can be solved independently. Then the local solutions are combined in order to obtain a feasible solution of the initial instance. Herein the method is used to solve the Static Lightpath Establishment problem (SLE), but can be easily applied to any RWA problem. We validate the proposed method on a real network and on large random instances. The obtained results are compared with the optimal ones or with those of a sequential algorithm.