Modelo de otimização da localização de estações radiobase baseado em sistemas imunológicos artificiais.

Abstract

This work describes base station deployment as a multi-objective problem (MOP). Base station deployment and configuration involve a large number of variables and design constraints, and heuristic algorithms seem to be a suitable alternative to solve MOPs. A new class of evolutionary algorithms, the so-called multi-objective optimisation algorithms based on artificial immune systems, are the basis of an innovative approach to base station placement. The Binary-coded Multi-objective Optimisation Algorithm (BRMOA) is presented. Two different scenarios are considered. In the first scenario, candidate sites have equal deployment costs and three network simulation environments are used for analysis. In the second scenario, cost-effective base station deployment is considered. The cost of deployment of a site is estimated based on its location and the Brazilian environmental legislation. The model allows some experience to be added in order to guide the process and lead the search to previously selected sites. Two different case studies are examined. Results are compared to the literature and indicate the feasibility of the optimisation strategy in solving a base station deployment problem.