Neuromodeling for microwave design

Due to its ability to learn and generalize from data and the nonlinear processing nature, neural networks are very suitable for modeling in multidimensional nonlinear problems, in the case of microwave design. In this paper, in order to confirm this fact, a brief review of neural network applications in different areas of microwave engineering is shown. Also, novel concepts, which incorporate microwave empirical knowledge for more efficient neural network modeling, are presented.