Microwave absorbers

Abstract

This research presents a novel polarization-insensitive, angular stable, tunable fractal frequency selective surface (FSS)-based GigaHertz (GHz) multiband and dualband absorber. A Neural network (NN) approach based is proposed and used to optimise the absorber to the desired frequency and bandwidth. The smallest mean square error for multiband absorber between the target and observed outputs is 0.015214603, and for the boradband absorber the smallest mean square error between the target and observed outputs is 0.0004774. Hence more close the mean square error is to the zero, more better the NN model is. At a central frequency of 10.5 GHz, maximum absorption of 62 percent is achieved for multiband absorber. Where as at a central frequency of 10.5 GHz, maximum absorption of 85 percent is achieved for broadband absorber. Furthermore, the designed absorber works well for both perpendicular and parallel polarizations with incidence angles up to 450 . The results from the NN were confirmed with the ansys software, and there was around 98% accuracy in NN to predict the frequency. THz imaging, sensing, detection, and stealth are all possible applications for the proposed absorber.