Low Profile Dual Band H-Slotted DGS Based Antenna Design Using ANN for K/Ku Band Applications

Abstract

In this manuscripts, H-slotted DGS based antenna configuration for K/Ku applications is designed, fabricated and measured using ROHDE & SCHWARZ ZVB20 network analyzer. Simulation studies are performed using Electromagnetic Simulation Software CST Studio Suite. Moreover, a development of Artificial Neural Networks (ANN) model is implemented based on feed-forward back-propagation and Bayesian regularization learning algorithm. The optimal position for H-slotted DGS antenna performance is predicted in terms of reflection coefficient and resonance frequency. The antenna is implemented on Rogers RT/Duroid 5880 with relative dielectric constant of 0.0009 and thickness 0.38 mm. It operates at both 15.04 GHz (from 14.87 GHz to 15.208 GHz) and 24.77 GHz (from 24.404 GHz to 25.137 GHz). The antenna achieves gains of 5.59 dB and 6.10 dB and efficiencies of 87 % and 80 % at both frequencies, respectively. The simulation using CST, predicted ANN and measurement results are in good agreement.