Otimização de redes de fibra óptica usando algoritmos de busca discreta cuco híbridos com algoritmo genético

• Ano de defesa: 2022
• Autor(a) principal: Lima, Rayane Araujo
• Orientador(a): Carneiro, Marcos Lajovic
• Banca de defesa: Dantas, Maria José Pereira , Vieira, Flávio Henrique Teles
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Pontifícia Universidade Católica de Goiás
• Programa de Pós-Graduação: Programa de Pós-Graduação STRICTO SENSU em Engenharia de Produção e Sistemas
• Departamento: Escola Politécnica
• País: Brasil
• Palavras-chave em Português: Problemas combinatoriais; Otimização de rede; Algoritmo busca cuco; Algoritmo busca cuco híbrido.
• Área do conhecimento CNPq: Engenharias; Engenharia de Produção
• Link de acesso: http://tede2.pucgoias.edu.br:8080/handle/tede/5074

Planning an optical network represents one of the most challenging in the telecommunications industry. Its discrete and combinatorial nature presents technical and economic problems, such as the resources¿ allocation, links and nodes, protection from failures, budget limitations, in addition to being costly for designers and engineers, among others. This work proposes some strategies to optimally design OTN/WDM networks. The network can be considered as a graph that has dynamic data traffic to flow from the source nodes to the destination. The proposed strategies deal with one of the main elements that affect the cost-performance ratio of an optical network: the design of the physical network topology (PTD). For the PTD problem, capital (CapEx) and operational (OpEx) costs are considered for analysis of network performance. Two models of metaheuristic algorithms are proposed, the first uses an adaptation of the cuckoo search algorithm and the second uses the hybrid cuckoo search algorithm with the genetic algorithm. these models are used to solve the MILP problem, for the planning of the optical transport network (OTN), considering cost minimization. Restrictions on guaranteeing demand fulfillment, specificity of the equipment technical capacity and exclusivity in the allocation of transmission link modularity are also contemplated. Therefore, the goal is to provide strategies that enable decision-making efficiently and realistically, taking into account the application scenario that is intended to portray. The computational results found are promising, the optimization resulted in topology models with reduced costs and relatively low processing time for a large network