Particle Swarm Optimization for base station placement

This article shows the improvement of using Particle Swarm Optimization (PSO) for multiple Base Station (BS) placement in a metropolitan area. We evaluate the algorithm´s performance using a combination of Shannon´s capacity formula and Jain´s index of fairness for two sets of traffic demand points, corresponding to an estimation of average and peak traffic, respectively. We show results performed by using 8, 32, 128 and 256 particles to place sets of new BSs versus number of iterations. We also exhibit potential optimal points for placement found by PSO. The optimization improved the average capacity by 17% with an increase on the number of BSs smaller than 10%.