Analysis of Parameter Settings for Differential Evolution Algorithm to Solve a Real-World Frequency Assignment Problem in GSM Networks

摘要

Frequency assignment is a very important real-world problem, specially in GSM networks. These networks are very used in the telecommunication area (by mid 2006 GSM services were used by more than 1.8 billion subscribers across 210 countries, representing approximately 77% of the world's cellular market). In this paper we solve a real-world instance of this problem, using a Differential Evolution (DE) algorithm hybridized with a Local Search method. We also analyze the performance of the several configuration parameters, because the performance of optimization algorithms is highly dependent on the specific properties of the problem to be solved. Several experiments were carried out to find the best set of parameters for the DE algorithm implemented in this work. The final results obtained by DE are very good.