During recent years, the artificial neural network (ANN) has been applied t
o electromagnetic (EM) engineering as a fast and flexible method for modeli
ng, simulation, and optimization. In this paper, an ANN model for the frequ
ency dependence of a radar cross section (RCS) is proposed and trained by m
eans of the measured data. Using this trained ANN model, the frequency depe
ndence of an RCS can be extended to a lower or higher frequency region. The
extended results are in reasonably good agreement with the measured and co
mputed results by the optical model published. (C) 1999 John Wiley & Sons,
Inc.