Canadian Journal of Pure & Applied Sciences

Abstract: Traffic grooming, which is the combination of traffic demands into a single wavelength channel is a well-known issue in Wavelength Division Multiplexing (WDM) optical networks. Grooming allows wavelength channels with high transmission capacity to serve many low-rate traffic demands simultaneously. In this paper, we address the traffic grooming, routing and wavelength assignment (GRWA) problem for WDM optical networks by considering multiple design objectives: maximizing the number of demands (commodities) served, minimizing the number of wavelength channels assigned, and minimizing number of transmission ports required. We use a hybrid multi-objective evolutionary computation approach consisting of Genetic Algorithm for routing allocation, Extended Traffic Grouping for traffic grooming and Maximum Degree First for wavelength assignment (GA-ETG-MaxDF). Then we apply the Fast Nondominated Sorting Genetic Algorithm (NSGA-II) to search for the set of non-dominated candidate solutions in multiobjective space. We compare the simulation results obtained from our approach (GA-ETG-MaxDF) with the alternative approaches (MST and MRU) published in the literature. We also examine standard performance metrics for multiobjective optimization solutions such as Hyper-volume, Spread, and Inverted Generational Distance. Based on our results, we conclude that the proposed technique is effective for solving the multi-objective GRWA problem in WDM optical networks.