Routing and wavelength assignment in all-optical networks
IEEE/ACM Transactions on Networking (TON)
Journal of Global Optimization
Computer Networks: The International Journal of Computer and Telecommunications Networking
Optical Networks: A Practical Perspective, 3rd Edition
Optical Networks: A Practical Perspective, 3rd Edition
Differential Evolution: A Survey of the State-of-the-Art
IEEE Transactions on Evolutionary Computation
Computer Networks: The International Journal of Computer and Telecommunications Networking
IEEE Journal on Selected Areas in Communications
Network Dimensioning under Scheduled and Random Lightpath Demands in All-Optical WDM Networks
IEEE Journal on Selected Areas in Communications - Part Supplement
Hi-index | 0.00 |
The routing and wavelength assignment (RWA) problem, known to be an NP-complete problem, seeks to optimally establish routes and adequate wavelengths for the requested connections according to an objective function. This paper presents the use of a novel approach based on a differential evolution (DE) algorithm to the RWA problem in wavelength-routed dense division multiplexing (DWDM) optical networks. The proposed DE-RWA algorithm is modeled to optimize not only the network wavelength requirement ($$ NWR $$NWR, which is the minimum number of wavelengths needed to fulfill traffic demand) but also the average path length ($$ APL $$APL). We present the impact of the control parameters of the DE algorithm on the improvement of system's performance. Additionally, we present two strategies to improve the efficiency of the algorithm, knowing as the disjoint cut-set paths (DCS-P) algorithm and the use of a random mutation ($$ random -M$$random-M) parameter for DE. The proposed approach is evaluated for test bench optical networks with up to 40 nodes. Experiments show that the DE-RWA algorithm obtains results that equal the $$ NWR $$NWR lower bound for networks with and without wavelength conversion capability, whereas reduce the $$ APL $$APL. The performance of the DE-based approach is compared against results obtained with the particle swarm optimization (PSO) and genetic algorithm (GA) models, showing that the DE-RWA outperform those algorithms. The presented DE-RWA model is simple to implement and could also be extended by adding other features such as impairment-aware, energy efficient, survivability among others in optical networks.