Optimum Positioning of Base Station for Cellular Service Devices Using Discrete Knowledge Model

Improved Multiobjective Optimization for Cellular Base Stations Positioning

“…The optimization of the tower positioning process begins similarly to the one proposed in [3], in which the land is discrete and classified according to building and vegetation criteria. The data used were those acquired in [3] at the frequencies of 2.1 GHz and 2.6 GHz, as well as the loss model proposed in that work.…”

“…In this work, in order to improve the study [3], the following complementary criteria were adopted for losses by type of environment: (i) a priority metric (based on the flow of people) by type of environment; (ii) a division of the study space in two parts. A reduction of up to 72% in processing time was measured between calculations without and with criterion (ii).…”

“…A reduction of up to 72% in processing time was measured between calculations without and with criterion (ii). A flowchart of the methodology proposed in [3] and the complementations proposed in this study are shown in Fig. 1.…”

“…Therefore, the studied scenario (the UFPA campus in Belém) was partitioned into two large partitions. This is one of the differences of this study, when compared to work [3]. This partition was made so that the number of candidate sites in each of the partitions was similar, this fact is important for the optimization process described in item V of this article.…”

“…This article proposes improvements in the work described in [3], which presents a model based on machine learning applied to a tool developed for the optimal positioning of Radio Base Stations (RBS), on the UFPA campus in the city of Belém, state of Pará, Brazil. The main objective is to improve the results obtained at the time, making them closer to reality in an eventual coverage project in the region considered.…”