Utilizing evolutionary algorithms for multiobjective problems in traffic engineering

Authors

S. C. Erbas,

Abstract

        In this study, we focus on a multiobjective mixed-integer off-line routing problem for the QoS traffic in Multi-Protocol Label Switching (MPLS) Networks. The methods to locate the nondominated front of a multiobjective optimization problem can be exact or approximate. Previously, we have investigated one of the exact methods, based on the aggregation of multiple objectives into a single function and iterative solution of these single objective problems. However, our experience with exact methods has proven that heuristic methods should also be investigated, as they may be favorable due to their speed and efficiency. In a case study, we have solved the same routing problem using a hybrid heuristic method which combines an evolutionary algorithm with mathematical programming, and compared the solutions obtained against those found by the exact method.

BibTEX Reference Entry 

@inproceedings{Er03,
	author = {Selin Cerav Erbas},
	title = "Utilizing evolutionary algorithms for multiobjective problems in traffic engineering",
	booktitle = "INOC 2003",
	address = {Evry/Paris},
	month = Oct,
	year = 2003,
	hsb = RWTH-CONV-223544,
	}

Downloads

 Download bibtex-file

Sorry, this paper is currently not available for download.