Unconventional Control of Electromagnetic Waves with Applications in Electrically Small Antennas, Nondiffracting Waves, and Metasurfaces

Abstract

Although electromagnetics is a well-established field within physics and engineering, it is also a rather dynamic one. Our ever increasing need for connectivity and the rise of medical and military applications constantly create new challenges for researchers in electromagnetics. In this thesis, unconventional methods are proposed for tackling some of these challenges by manipulating the electromagnetic fields in different regions: reactive near field, radiative near field, and far field. The first topic examined pertains to the development of antennas for Internet of Things (IoT) nodes with extremely small form factors and low power consumption. As a result, they require small and relatively efficient antennas that can be tightly integrated within the node. The antennas should be able to operate with lossy and metallic components in their near field, while maintaining adequate performance. A type of antenna, called a 3D loop, is designed to fulfill these specifications, and is used in two compact IoT systems. The second thrust aims at developing devices that generate Bessel beams and X waves in their radiative near field. Bessel beams are a class of exotic beams with nondiffracting and self-healing properties. Here, two radiator designs are presented, capable of generating Bessel beams with minimal deviation of their parameters over a broad bandwidth. This allows the generation of nondiffracting and nondispersive pulses (X waves) that remain highly localized within the device’s radiative near field. The final topic examined aims at analytically modeling the electromagnetic properties of patterned metallic sheets. Such sheets are the building blocks of metasurfaces, which are two dimensional devices that manipulate the properties of a propagating wavefront (amplitude, phase, polarization). Having analytical models for sheets that realize arbitrary electromagnetic properties significantly expedites their design, as opposed to relying on databases of simulated geometries.