One of the most important issues in optical network design is ensuring its resilience. In this paper, we propose using nature-inspired heuristics to find a resilient mapping of a given virtual topology with minimum resource usage. Evolutionary algorithms and ant colony optimization algorithms are applied to the problem after a set of parameter tuning tests. To assess the performance of the proposed algorithms, we compare the experimental results with those obtained through integer linear programming. The results show that both of our algorithms can solve the problem even for large-scale network topologies for which a feasible solution cannot be found using integer linear programming. Moreover, the CPU time and the memory used by the nature-inspired heuristics is much lower. The solution quality and the CPU time usage results prove that both of our nature-inspired heuristics can easily be applied to real-world applications.
© 2010 Optical Society of AmericaFull Article | PDF Article
Ignacio de Miguel, Reinaldo Vallejos, Alejandra Beghelli, and Ramón J. Durán
J. Opt. Commun. Netw. 1(7) 608-621 (2009)
Rui Manuel Morais, Claunir Pavan, Armando Nolasco Pinto, and Cristina Requejo
J. Opt. Commun. Netw. 3(1) 17-26 (2011)
Oscar Pedrola, Davide Careglio, Miroslaw Klinkowski, and Josep Solé-Pareta
J. Opt. Commun. Netw. 3(9) 651-666 (2011)