A memetic algorithm approach for the planning and optimization of a new-generation cellular network capitalizing on existing sites

Abstract

In a context of technology migration, controlling the costs of deploying a new cellular network is essential. Several algorithmic approaches have proven themselves in the field of telecommunications for optimizing the placement of sites in an operator infrastructure. The work carried out in this article evaluates the effectiveness of a memetic approach in minimizing the cost of deploying the next generation cellular network. We this models the cellular network migration problem as a question of optimization, which we resolve using this approach resulting from the combination of a genetic algorithm and a taboo search. Using the Jupyter tool, we created a model that takes as input an area in which a set of operational sites and potential locations of new sites are deployed. We then implement the genetic algorithm and then associate it with a taboo search based on the minimization of new sites and the reuse of existing sites. During testing, the proposed memetic approach uses 31 % of all sites; while the genetic algorithm alone uses 33%. In addition, we observe an increase in the coverage rate which goes from 76.1% to 86.4% with the memetic approach.