Pareto based bat algorithm for multi objectives multiple constraints optimization in GMPLS networks

Masood, Mohsin and Fouad, Mohamed Mostafa and Glesk, Ivan; Hassanien, Aboul Ella and Tolba, Mohamed F. and Elhoseny, Mohamed and Mostafa, Mohamed, eds. (2018) Pareto based bat algorithm for multi objectives multiple constraints optimization in GMPLS networks. In: The International Conference on Advanced Machine Learning Technologies and Applications (AMLTA2018). Advances in Intelligent Systems and Computing . Springer, EGY, pp. 33-41. ISBN 9783319746906 (

[thumbnail of Masood-etal-AMLTA-2018-Pareto-based-bat-algorithm-for-multi-objectives-multiple-constraints-optimization]
Text. Filename: Masood_etal_AMLTA_2018_Pareto_based_bat_algorithm_for_multi_objectives_multiple_constraints_optimization.pdf
Accepted Author Manuscript

Download (793kB)| Preview


Modern communication networks offer advance and diverse applications, which require huge usage of network resources while providing quality of services to the users. Advance communication is based on multiple switched networks that cannot be handle by traditional IP (internet protocol) networks. GMPLS (Generalized multiprotocol label switched) networks, an advance version of MPLS (multiprotocol label switched networks), are introduced for multiple switched networks. Traffic engineering in GMPLS networks ensures traffic movement on multiple paths. Optimal path(s) computation can be dependent on multiple objectives with multiple constraints. From optimization prospective, it is an NP (non-deterministic polynomial-time) hard optimization problem, to compute optimal paths based on multiple objectives having multiple constraints. The paper proposed a metaheuristic Pareto based Bat algorithm, which uses two objective functions; routing costs and load balancing costs to compute the optimal path(s) as an optimal solution for traffic engineering in MPLS/GMPLS networks. The proposed algorithm has implemented on different number of nodes in MPLS/GMPLS networks, to analysis the algorithm performance.