Diseño y optimización genética de antenas pixeladas reconfigurables

Abstract

Historically, antennas have been considered as static elements in communication systems. However, during the last ten years, the concept of recon gurable antenna has gained importance. This concept consists on an antenna which can modify its properties by mechanical, electrical or thermal actuation, obtaining a better adaptation to the communication channel and making possible the integration of several antennas in a single recon gurable antenna. Some of the structures with a higher recon guration capability are pixelled geometries, which consist of a grid of metallic patches linked by switches. These structures use optimization algorithms to choose the state of the switches, conveniently modifying this way the current distribution on the antenna surface in favor of a certain goal. In this project the di erent alternatives for the design and optimization of pixelled antennas have been analyzed. This analysis has been used to develop three di erent designs following several architectures: monopole, dipole and patch. Electromagnetic simulations had been carried out for these cases to demonstrate their recon guration capability. Additionally, two prototypes of the monopole design have been manufactured and measured with satisfactory results. One of the main contributions of this work has been the development of a hybrid pixelled geometry using patches of di erent sizes with the aim of reducing the structure complexity and the number of necessary switches. Another contribution has been the design of an antenna which achieves recon gurability using an external pixelled layer. This is a novel idea because it makes possible to provide basic antennas with recon gurability by means of a complete external way avoiding the modi cation of the initial design.