Adaptive neuro fuzzy inference system for the optimization of behavior of fractal antenna

This paper presents an Adaptive neuro-fuzzy inference system (ANFIS) that helps to approximate the characteristics of an Ultra wideband fractal antenna, quickly and with more accuracy. The optimization of an antenna with respect to various parameters using full wave analysis techniques takes more computational time, whereas, the simulations conducted by the author reveal that, a trained ANFIS based on Least Square Estimator and Back propagation gradient descent algorithm surpasses the time-tedious repetitive simulations, by approximating its own behavior, and makes the calculation easy and swift. The simulated results of High Frequency Structure Simulator (HFSS) are employed in creating an optimized model of a fractal antenna using ANFIS. After training ANFIS model, the simulated results obtained are compared with the results of HFSS, and they are found to be in close agreement within the deviation of 1.5525%. Thus ANFIS is found to be an effective alternate to full wave analysis in Optimizing the Impedance Bandwidth and radiation pattern of Antenna.