Multi-Space Competitive DGA for Model Selection and its Application to Localization of Multiple Signal Sources

In this paper, a new optimization method, which is effective for the problems that the optimum solution should be searched in several solution spaces, is proposed. The proposed method is an extension of Distributed Genetic Algorithm (DGA), in which each subpopulation searches a solution in the corresponding solution space. Through the competition between the sub-populations, population sizes are adequately and gradually changed. By the change of the population size, the appropriate sub-population attracts many individuals. The changing population size yield the efficient search for the problems of searching for solutions in multiple spaces. In order to evaluate the proposed method, it is applied to a polynomial curve fitting and signal source localization, in which the number of sources is preliminarily unknown. Simulation results show the effectiveness of the proposed method.