A Genetic Approach for the Efficient Numerical Analysis of Microwave Circuits

摘要

The development of effiecient and effective algoritms for sparse matrix bandwidth minimization is of paramount importance for the enhancement of many numerical techinques for the analysis of microwave circuits. The tast of bandwidth reduction is computationally hard. Several approaches have already been proposed, but the problem is still open. In this paper, a genetic solution is proposed. The genetic algorithm is described, as well as its main characteristics (choice of chromosomes, genetic operations, etc.). Results demonstrate that the advantages of the genetic approach vanish because of the huge computational effort required. This severe limitation is removed thanks to the natural amenability of genetic algorithms to a parallel implementation. Results in the paper prove that a parallel genetic approach is a state-of-the-art solution to the problem of bandwidth reduction of sparse matrices encountered in electromagnetic numerical methods.