Multi-objective Optimization in Optical Networks

Optical networks use very effectively the available fiber bandwidth and thus provide the technological platform for core networks supporting various services, e.g. 4G and 5G mobile systems. From the network operator point of view, the most important element of the optical network design process is the minimization of network resources and estimate the moment of congestion in terms of maximum network capacity. Hence, the main objective of this study is firstly the minimization of optical nodal resources, i.e. transponders and amplifiers. Consequently, the optimization problem we consider is multi-objective. Heuristic methods based on the evolutionary algorithm were proposed to solve the problem, which was compared with the reference algorithm. The results obtained show that there is an advantage in using heuristic algorithms for optimizing large optical networks. The numerical experiments were carried out for the network, which is typical of a realistic network.