Proposta de rede GPON usando PSO

Abstract

This study presents the proposes the use of Particle Swarm Optimization (PSO) meta-heuristic, in the process of building a Gigabit Passive Optical Network (GPON) with Fiber To The Home (FTTH) architecture. The FTTH network based on the GPON standard is considered a technology option capable of disseminating high-speed Internet, however, the high cost of the project may make its implementation unfeasible. Faced with this challenge, the PSO is applied to reduce the cost of the network, given that the planning of an optical network involves the geographical determination of the positioning of different equipments, fiber optic cables, and several other parameters relevant to the proper functioning of the network, being an arduous and complex task. In summary, we develop a PSO algorithm in Matlab® using the PSO, which allows us to quickly and efficiently develop solutions for GPON projects using real maps with experimental data, as well as minimizing implementation costs and reducing the time allocated for the elaboration of projects. The efficiency of the proposed computational system was confirmed from comparisons with PONs networks elaborated by manual and random creation techniques. Based on the obtained results, it was proved that the PSO based network design has a better average cost, bringing viable and attractive results, in addition to an acceptable execution time, attesting to the potential of the system in the proposal of GPON networks.