Brinhole, E. R.Destro, J. F. Z.Freitas, A. A. C. deAlcantara, N. P. de [UNESP]2014-05-272014-05-272005-12-01PIERS 2005 - Progress in Electromagnetics Research Symposium, Proceedings, p. 579-582.http://hdl.handle.net/11449/68592This paper presents models that can be used in the design of microstrip antennas for mobile communications. The antennas can be triangular or rectangular. The presented models are compared with deterministic and empirical models based on artificial neural networks (ANN) presented in the literature. The models are based on Perceptron Multilayer (PML) and Radial Basis Function (RBF) ANN. RBF based models presented the best results. Also, the models can be embedded in CAD systems, in order to design microstrip antennas for mobile communications.579-582engAntennasBackpropagationComputer aided designEmbedded systemsFeedforward neural networksMicrostrip antennasMicrowave antennasMobile telecommunication systemsNatural frequenciesNeural networksPiersRadial basis function networksWireless networksArtificial neural networksCad systemsEmpirical modelsMobile communicationsPerceptronRadial basis functionsResonant frequenciesMobile antennasTrabalho apresentado em evento10.2529/PIERS041210091305Acesso aberto2-s2.0-55749095422