Field Manipulating Surfaces

5G and 6G communication networks aim to ensure fast data rates, wide bandwidth, improved coverage and minimal latency; operating in mm-wave and sub-THz frequency bands, however, presents remarkable challenges including free-space loss, building penetration loss and strong interactions with obstacles, possibly leading to coverage gaps. Integrating active and passive devices into the environment has been widely proposed as a means of enhancing coverage without expanding the number of base stations. Among the others, Smart Electromagnetic Skins (SESs) offer a passive, thin-surface solution for redirecting incident fields in desired directions; depending on specific applications, SESs may serve as either reflective or transmitting devices, contributing to the creation of a Smart Radio Environment (SRE). In particular, a passive transmitting SES integrated into windows presents an opportunity to strengthen the connection between base stations and indoor terminals. However, due to its specific placement, the design must balance performance with minimal interference in visibility and light passage. Traditional structures consist of meshed opaque conductors on transparent substrates but often neglect window integration. Alternative designs have introduced multi-layer structures directly embedded in the window, printing a thin conductive layer between two glass panes. Since metallization affects both transparency and radiation efficiency, recent research efforts have explored the potential of a fully dielectric transparent smart skin that is directly integrated with the window. This is the solution investigated in this Thesis: in particular, an innovative Unit Cell (UC) model is introduced, focusing on the analytical design of multi-layer, purely dielectric configurations. This design method is presented and validated in a variety of test cases, in order to provide a tool that can be exploited at design time without the need to resort to CAD simulations, which are both time-consuming and computationally heavy. The strengths of the proposed scheme are represented not only by its speed and its more-than-satisfactory level of accuracy, but also by its suitability for the characterization of complex multi-layer cells. Additionally, the method's adaptability extends to tapered structures, modeled as non-uniform transmission lines, further broadening the potential applications. In order to enhance the performance of the UC while fulfilling the phase constraint required when the cell is embedded in a transmitting surface, an optimization strategy based on the Genetic Algorithm is developed, with emphasis on reducing back radiation and control of the thickness of the entire structure, including the glass layer. Preliminary findings, presented for a number of geometric configurations, show that, accepting a small phase error, the performance of the cell can be significantly improved, obtaining satisfactory values for both reflection and transmission coefficient in most angle configurations. The approach used for the UC is then extended to transmitarray and SES design and validated through full-wave electromagnetic simulations, demonstrating the model's effectiveness for both feed and plane-wave incidence. Ultimately, this analysis offers a faster alternative to typical simulation-based design while significantly reducing computational costs, paving the way for broader exploration of novel configurations in future research.