Flow-based path selection for Internet traffic engineering with NSGA-II

Traffic engineering has become an important issue in Internet operation due to the fast growing of the Internet traffic and the stringent requirements of quality-of-service over the limited available resources. This problem is a multicriteria optimization problem in nature and our goal in this paper is to explore the application of NSGA-II, an evolutionary algorithm for multiobjective optimization, for determining the optimal distribution of traffic demands over the network. The problem is first formulated as a multiobjective constrained optimization problem which is NP-hard. Then, a hybrid heuristic algorithm based on of linear programming and NSGA-II is developed for approximating the optimal Pareto front. We compare the performance of the proposed heuristic using a 10-node problem adopted from the literature with the exact solutions generated using a lexicographic Chebyshev method.