Soft Computing based Ellipsoidal Fractal Antenna for Wide Band Applications

The paper proposes soft computing based neural network model for analysis of parameters of ellipsoidal fractal microstrip antenna (EFM Antenna) utilizing wide band applications. Sierpinski Gasket fractal concept is used to design the radiating patch with second degree of iterations. The design parameters like minor axis, major axis of ellipse, height of substrate, number of iterations, resonant frequencies, and bandwidth are used for the data collection in designing the model for predicting the efficient values in the design of the proposed antenna. Artificial neural network model is used here for training, testing and validating the sampled data for analysis. The model is trained with 75 data sets and 25 sets for testing purpose. The algorithm used here for this purpose is multilayer perceptron. The result of soft computing model is found in close agreement with the simulated results. Also, the minor and major axis of ellipse for the second iteration of fractals suggest the wide band and multiband operation of the antenna in Ku (12-18 GHz) and K(18-27 GHz) band applications.