Antenna Optimization Using Adaptive Neuro-Fuzzy Inference System (ANFIS)

Abstract

Owing to the increasing demands of miniature, cost-effective, light weight, low profile and wideband antenna, the concept of fractal antenna came into light.The prime advantage associated with fractals is its multiband operations. However, the mathematical analysis of the fractal antenna is difficult due to its complex shape and geometries.Along with this, it requires user expertise, more memory and a large amount of time to design the antenna using full-wave solvers such as High Frequency Structural Simulator (HFSS).In that case Adaptive Neuro-Fuzzy Inference System (ANFIS) is an alternative approach which is easier to implement and provides a good solution with minimum resources in lesser time. The presented work aims at investigating the application of the adaptive neuro-fuzzy model for the analysis and synthesis of various fractal antennas suitable for wireless applications.A compact, multiband antenna based on the Sierpinski Carpet Fractal is designed to operate at 0.5 to 3 GHz and a compact antenna based on Sierpinski Gasket Fractal is designed to operate from 1 to 7 GHz. For analysis model of Sierpinski Carpet Fractal Antenna(SCFA) and Sierpinski Gasket Fractal Antenna(SGFA),the geometrical parameters of the antenna are considered as the input and the resonant frequencies are taken as the output while a reverse process is followed for the synthesis model. Also, a model is trained to obtain the return loss characteristics for the given set of geometrical parameters. Finally the antenna is simulated in order to predict the radiation pattern characteristics. The results obtained through the proposed model are used to simulate the fractal antenna.The output of all the proposed models are compared with that of HFSS and both of them are found in close agreement.Lastly the Minkowski Fractal antenna was simulated by using the HFSS.