Neural network modeling of scattering parameters from a conducting post in rectangular waveguide

Neural networks are popular as efficient alternatives to conventional computational models like numerical modeling or analytical methods for RF and Microwave modeling and design. Such models reduce both processing time and analytical complexity resulting in simpler computation process. Thus the technique has greater potential in the analysis of various problems in electromagnetism. In the present problem neural network technique has been used for non- linear modeling of the frequency response of the scattering properties of a long circular conducting cylinder in a three dimensional rectangular waveguide at X band. The scattering properties of the vertical posts are analyzed by semi analytical approach based on image theory and the scattered fields are calculated from the lattice sum and the transition matrices. Analysis results have been used to generate S parameter data for the configuration in order to train the proposed network.