Cellular Radio Network Planning using Particle Swarm Optimization

The cell planning problem concerns finding a set of sites for antennas from a set of pre-defined candidate sites, and for each selected site, to determine the number of antennas, as well as the associated values for each of the antenna parameters. All these choices must optimize a set of objectives. This paper presents a new approach for tackling this complex and highly combinatorial problem. The proposed approach is based on the application of the powerful optimization technique known as particle swarm optimization (PSO). First, the working area is represented by a suitable model showing topological and morphological information and the places for the candidate sites. After that, a solution representation is proposed converting each network plan to a bit stream. Finally, we apply the PSO algorithm to reach the best network plan satisfying our objectives. To validate the approach, computational results are presented using realistic data designing a cellular radio network in a part of Libreville, the capital of Gabon.