Differential evolution optimization applied to the wavelength converters placement problem in all optical networks

The placement of wavelength converters in an arbitrary mesh network is known to belong to the class of NP-complete problems. So far, this problem has been solved by heuristic strategies or by the application of optimization tools such as genetic algorithms (GAs). In this paper we introduce the application of Differential Evolution (DE) to the problem of the placement of wavelength converters to obtain the optimal solution. Many comparative studies confirm its robustness and efficiency, showing that in many cases DE outperforms many other well known evolutionary computational approaches in terms of convergence speed and quality of solutions. The major advantage of the DE algorithm rests in the fact that it does not need to build up a search tree or to create auxiliary graphs as other methods do. Furthermore, the method typically requires few control parameters, and the computed results show that only a small population is needed to obtain the optimal solution for the placement of wavelength converters in an arbitrary network. We present experiments that demonstrate the effectiveness and efficiency of the proposed evolutionary algorithm.