Material Perturbation in Rectangular Dielectric Resonator Antenna Using Neural Network

Abstract

The paper presents a concept of material perturbation in rectangular dielectric resonator antenna (RDRA), which can be helpful for gain and bandwidth enhancement over original DRA. Material perturbation provides a wide tuning range, improved degree of freedom, and hence more flexibility in designing application-specific antennas. The perturbation parameters are optimized using artificial neural networks for a given design frequency. The concept of Perturbation was previously used as layered and stacked DRAs. Further, material perturbation and wall perturbation can be combined for obtaining polarization diversity in DRAs.