| dc.description.abstract | Studies on network infrastructure have been used and applied in various service industries, such as telecommunications, power distribution, water, and more recently to meet the smart city infrastructure needs, that require hybrid infrastructure solutions. However, planning multilevel network infrastructures and also with maximum coverage is an open problem, since the literature generally presents solutions in which only one level is processed and problems solved individually. The distribution plan and equipment connection at various levels of a network infrastructure is a ardous task, it is necessary to evaluate the quantity and better geographical distribution of equipment at each level of the network, ie at level 1 are evaluated needs to meet directly to customers, since the level 2 from level 1 will meet the needs, and the adjacent levels are also observed. The result that a designer gets to manually carry out the planning becomes shallow due to impossibility of analyzing all scenarious that can be obtained with the amounts and distribution of information equipment in the project region. Even more complex is change
a level already designed, since it is necessary to review the amount, location and path between devices, redoing the calculations of the modified level, and hence the previous level and all subsequent levels. This research presents an algorithm based on ant colony metaheuristics, capable of searching for a solution to the infrastructure planning of hybrid multilevel networks, solving the network planning and coverage problem together and obtaining better results in the order of 20% on the cost of the solution when compared to another solution using genetic algorithm. | en |
