Algoritmo genético especializado aplicado à otimização do planejamento de redes óticas passivas

• Ano de defesa: 2020
• Autor(a) principal: Pereira, Oeber Izidoro
• Orientador(a): Franco, Edgar Manuel Carreño
• Banca de defesa: Freitas, Ricardo Luiz Barros de , Pareja, Luis Alfonso Gallego
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual do Oeste do Paraná; Foz do Iguaçu
• Programa de Pós-Graduação: Programa de Pós-Graduação em Engenharia Elétrica e Computação
• Departamento: Centro de Engenharias e Ciências Exatas
• País: Brasil
• Palavras-chave em Português: Algoritmo genético de Chu-Beasley; Pesquisa operacional e otimização; Otimização de projetos de redes de PON; Redes de acesso de telecomunicações
• Palavras-chave em Inglês: Chu-Beasley genetic algorithm; Operational research and optimization; Optimization of PON network projects; Telecommunications access networks
• Área do conhecimento CNPq: ENGENHARIA ELETRICA::TELECOMUNICACOES
• Link de acesso: http://tede.unioeste.br/handle/tede/4943

The telecommunications services access networks have been the bottleneck for many users who want to use applications on demand, available over the Internet, such as: software as a service, storage, entertainment and social media. To meet this need, operators' access networks need to grow broaden and flexibly. Within this context, the Passive Optical Networks – PON technology, has proven to be a flexible, reliable and “future proof” option. However, the design of access networks for the delivery of general public services (water, electricity, gas and telecommunications) is a highly complex combinatorial problem, which hinders the global optimization of such projects. This work aims to optimize the design of PON networks, using the Chu & Beasley genetic algorithm, with two coding vectors applied to a georeferenced planning scenario, in order to achieve the lowest PON network deployment total cost. In the proposed algorithm, all relevant technology variables were consider, such as components and longest path attenuation, number of users per port (active and passive) and other relevant restrictive parameters. Another goal of this work was to identify and adjust the genetic operators that leads the algorithm to its best performance, improving the results already obtained on related work.