Traianos V. Yioultsis

We introduce a method of lines (MoL) formulation for analyzing and designing periodic multi-element millimeter-wave (mmWave) metasurfaces (MS) in a computationally tractable manner. As the complexity of MSs increases, their analysis and especially their design using full-wave methods becomes challenging. The MoL offers a considerably faster alternative, as it is a semi-analytical method where the electromagnetic equations are solved analytically along the direction perpendicular to the MS layers and numerically on the MS plane; hence the required degrees of freedom (DoFs) are substantially decreased. We employ an extended MoL formulation, where a first-order differential equation involving both the total tangential electric and magnetic field components is solved and the transfer matrix of each layer is obtained. As a proof of concept, we analyze various MS filter supercells, varying the polarization of the incident wave and the resonator parameters. Very good agreement with finite element method (FEM) simulations is observed in all cases. By comparing the extended MoL to the FEM in terms of computational requirements, we verify that the MoL presents a considerably faster alternative for MS design. The presented analysis paves the way for the fast synthesis of optimized MSs by changing only the boundary conditions on the layer interfaces. The advantages of our method will be even more pronounced in the design of even more complex devices.

The technology of reconfigurable intelligent surfaces (RISs) is widely expected to be a key factor in controlling the wireless communications environment. Due to their structural complexity, modeling RISs with simple but sufficiently accurate analytical models is mandatory for integrating these structures into new wireless 6G networks. We thereby propose a new, conceptually consistent free-space path loss (PL) model, based on firm electromagnetic (EM) theory principles as dictated by a more accurate physical optics (PO) approximation. In particular, we calculate the scattering properties of the unit cell, as an equivalent impedance sheet, using equivalent surface electric and magnetic current density via the PO model. Based on the EM analysis findings, we use the complex radar cross section (RCS) matrix to model each unit cell’s contribution to the overall scattering process. This allows accurate calculation of the scattering pattern and the received power considering polarization phenomena. Therefore, the developed free-space PL model is universal, as it incorporates RIS features to calculate scattering properties across both near- and far-field regimes and for both in-plane and out-of-plane reflection. A thorough comparative study is performed to validate the proposed analytical model, considering the RIS as an equivalent surface of discrete unit cells, using highly accurate 3D full-wave numerical simulations as a benchmark. Comparison of the proposed model to full wave simulations demonstrates its excellent accuracy and potential as a tool for highly accurate RIS modeling.

This paper explores graphene’s quasi-normal modes (QNMs) by developing a finite element method solver, which addresses an augmented eigenvalue problem where graphene is represented as an equivalent surface current. This representation correlates with the surface conductivity of the material through a Debye frequency dispersion model. Initially, the straightforward intraband term is considered, while the study delves into the interband contributions to graphene’s conductivity, described via a complex conjugate series of Debye terms. Also, it examines the magneto-optical properties of graphene, which can be tuned by an external magnetic field to produce non-reciprocal effects. Finally, the proposed formulation covers the normalization of graphene QNMs and the reconstruction of scattered fields, providing a complete analysis tool that handles graphene as a scattering surface. The featured methodology is successfully validated by comparing the evaluated absorption cross-section due to scattering from graphene sheets with the corresponding outcomes of a commercial full-wave solver. Graphene attributes are selected properly to model high quality-factor resonances indicating the remarkable accuracy of the proposed scheme.

We fully explore the properties of a novel uniplanar metamaterial-inspired substrate integrated waveguide (SIW) for realizing easy-to-fabricate mmWave components. Conventional SIWs have a complex fabrication process, especially for higher frequencies, due to the need for drilling holes and injecting metallization. We substitute the vias of the conventional SIW with grounded single complementary split-ring resonators (CSRRs). We employ a coupled-mode theory framework to guide the waveguide synthesis and gain insight into the underlying physics of periodic grounded uniform and multi-element CSRR metasurfaces, that are the key elements of the corresponding CSRR SIWs. The CSRR SIW parameters are then fine-tuned by a more detailed FEM eigenmode solver. We verify the uniplanar single-CSRR SIW’s merit and applicability by fabricating and experimentally characterizing the performance of a CSRR SIW waveguide section and antenna. The uniplanar single-CSRR SIW offers considerably simpler fabrication than both the conventional SIW and previously proposed CSRR SIW variations while maintaining performance. Compared to other SIW variations, the uniplanar single-CSRR SIW has a greater design flexibility, as we can change the number of CSRR rows and also use a non-uniform multi-element CSRR pattern to achieve advanced functionalities. The CSRR-SIW-based periodic leaky-wave antenna (PLWA) has competitive radiation characteristics and simpler fabrication than a corresponding SIW PLWA.

This study explores three-dimensional proactive beamforming at millimeter wave frequencies using transformer neural networks (TNNs), long short-term memory networks (LSTMs) and gated-recurrent units (GRUs). The proposed scheme aims to reduce beamforming latency by predicting future directions of arrival (DoAs) based on past observations, allowing the system to prepare beamforming weights proactively. We simulate an urban environment using OpenStreetMap data to generate realistic movement paths, creating a comprehensive dataset for training and evaluation. Our focus is on the predictive capacity of TNNs, LSTMs and GRUs to anticipate future DoAs, even in non-line-of-sight scenarios influenced by urban infrastructure. We detail the environment simulation setup, the ray-tracing mechanism as well as the movement generation process for pedestrians and vehicles. A statistical analysis on the prediction accuracy and response time is presented to assess the most accurate model and discuss the trade-offs between the architectures. In addition, an end-to-end AI-based proactive beamforming scenario is examined where zero-forcing is applied on moving users. This is to further demonstrate and evaluate the capabilities and the performance of each model. Our findings suggest that proactive beamforming can significantly enhance performance in dynamically changing urban landscapes, offering a promising avenue for future research and development in adaptive communication systems.

The performance of the metamaterial-inspired substrate-integrated waveguide is discussed in this work, concerning a resonator misalignment potentially caused by the fabrication process. Initially, the design parameters of the aforementioned waveguide at the X-band are presented, while its optimal operation is validated to prove the effectiveness of the apparatus. Then, various significant aspects of the polynomial chaos expansion theory are briefly introduced to facilitate the sensitivity analysis due to fabrication errors. The direction of misalignment is, firstly, investigated, while the general case is, also, considered, highlighting a notable immunity, especially at lower frequencies. Additionally, a parametric examination is conducted in terms of the fabrication tolerance, measured as a percentage of the resonator unit cell. All the required simulations are conducted utilizing the non-intrusive approach of the polynomial chaos methodology via the popular Finite-Difference Time-Domain scheme.

The functionality of a fully planar metamaterial-inspired substrate-integrated cavity is thoroughly investigated in the present work. The electromagnetic field confinement of the proposed device is realized with the utilization of broadside-coupled complementary split ring resonators that operate as a virtual electric wall. The numerical results from the eigenvalue analysis verify the presence of the fundamental resonances with a noteworthy quality factor. Subsequently, full-wave simulations are conducted to validate the resonance functionality of the proposed device, with excitation achieved using the metallic core of a coaxial cable. Numerical results highlighted, also, the radiation capabilities exploiting the inherent openings in the device due to the complementary resonators.

In the rapidly evolving domain of wireless networks, adaptive beamforming stands as a cornerstone for achieving higher data rates, enhanced network capacity, and reduced latency. This study introduces a novel integration of deep neural networks (NNs) into adaptive beamforming, specifically for uniform planar arrays (UPAs). We embark on an exploration of different NN architectures (a deep feedforward NN, a gated recurrent unit, and a long short-term memory based recurrent NN) and hyperparameter tuning techniques (grid search, Bayesian optimization, and genetic algorithm) to assemble a model that yields the most promising zero forcing beamforming performance. This thorough examination not only unveils the most prominent model but also highlights the most efficient hyperparameter tuning method. A comprehensive dataset generated using the null steering beamforming algorithm applied to an 8×8 UPA, serves as the bedrock for training these networks. Our evaluation benchmarks the NN-based beamformers against traditional methods, assessing them on crucial metrics such as beamforming competence, response time, and adaptability under various noise and interference scenarios. A significant innovation of this research is the introduction of a continual learning approach, enhancing the NN beamformers’ performance in dynamically changing environments. This self-improving algorithm represents a stride towards more adaptable and efficient beamforming, showcasing improvement on underperforming scenarios without compromising accuracy. Our findings demonstrate the potential of neural networks in reshaping the future of adaptive beamforming, offering a blend of speed and precision that is paramount in modern wireless networks.