Reconfigurable, miniaturized, and cognitive antennas for IOT devices.

Abstract

In recent years, the advancements in wireless communications imposed the need of an intelligent Wireless system conscious of its environment and capable of changing its operating frequency, modulation, waveform and transmitting power. Such advancements are accompanied by an imbalance use of the spectrum where some frequency bands are overloaded while others often remain idle. As a solution to the spectrum imbalance, cognitive radio enables sensing the spectrum, learning its activities and identifying the suitable frequencies for communication. Once such frequencies are identified, antennas are ordered to reconfigure their topology through software in order to achieve a successful communication across the desired bands. Cognitive radio becomes a necessity in the era of Internet of things (IoT), as millions of devices must be connected and able to communicate on demand. Enabling such devices to communicate requires the integration of agile reconfigurable antennas within their communication systems. Hence, such antenas must be miniature in size, reconfigurable and easily software controlled. At the same time these antenas must be able to preserve aceptable radiation efficiencies, stable radiation patterns and constant gains. In this tesis the challenge of proposing a new antena design that is miniature in size and exhibit an aceptable radiation performance is tackled. Furthermore, machine learning techniques come at the core of such design requirements by providing optimization tools that allow the successful miniaturization of the antena structure while satisfying all the required constraints. In fact, techniques such as genetic algorithm and quantum genetic algorithm can be employed to generate three dimensional and volumetric antena structures that can be integrated in compact IoT devices within a Dynamic cognitive radio setting. This tesis also addresses the integration of switching components within the antena topology in order to reconfigure its operation. Such integration benefits as well from artificial i