Galymzhan Nauryzbayev

Reconfigurable intelligent surfaces (RISs) are emerging as a key enabler of 6 G wireless systems, offering the potential to significantly enhance signal propagation and network performance. This paper presents a novel approach for modeling RIS-assisted millimeter-wave communication channels using a Gamma-distributed approximation. By modeling the channel gains between the base station, RIS, and user as Gamma random variables, the proposed method captures both line-of-sight conditions and shadowing effects effectively. We derive closed-form expressions for two critical performance metrics, such as outage probability (OP) and ergodic capacity, along with their corresponding upper and lower bounds. The analytical results are validated through extensive Monte Carlo simulations, demonstrating close agreement between the approximation and simulation data. Furthermore, we observe that the accuracy of the Gamma approximation improves as the number of RIS elements increases, and the OP bounds tighten under less severe channel conditions.

This paper introduces a unified optimization framework for bidirectional communication in simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) aided full-duplex (FD) systems. By leveraging unique characteristics of STAR-RIS modes: mode switching (MS), energy splitting (ES), and time switching (TS) - and accounting for practical channel state information (CSI) and successive interference cancellation (SIC) imperfections, we aim to optimize bidirectional multiple access for four operational regimes: 1) quality-of-service (QoS) feasibility, 2) energy efficiency (EE), 3) STAR-RIS efficiency, and 4) max-min fairness. For MS and ES modes, we employ the non-orthogonal multiple access (NOMA) protocol to derive a optimal STAR-RIS partitioning and energy splitting coefficients that optimizes performance subject to uplink (UL) QoS demands, then fine-tune base station (BS) power allocation to maximize downlink (DL) performance while ensuring compliance with DL-QoS. On the other hand, TS mode exploits the entire STAR-RIS for a single UE at a time, follows an orthogonal multiple access (OMA) approach and optimizes time allocation parameters individually for both UL and DL scenarios. Our performance analysis accounts for CSI/SIC imperfections. Extensive simulations confirm high consistency between analytical and numerical results, with SIC imperfections proving more disruptive due to heightened interference. While MS and ES outperform TS in terms of max-min rate under perfect SIC conditions (e.g., 5.57 bits/s/Hz in MS/ES modes compared to 3.3 bits/s/Hz in TS mode given N “ 128 and ϵ “ 0), TS yields superior results when SIC imperfections are present (e.g., 3.3 bits/s/Hz in TS mode compared to 2.05 bits/s/Hz in MS/ES modes given N “ 128 and ϵ “ 0.1).

UAVs are poised to play a pivotal role in next-generation communication networks, supporting applications from emergency coverage to large-scale IoT connectivity. However, enabling intelligent and autonomous UAV networking remains challenging because conventional Artificial Intelligence (AI) approaches have inherent limitations. Neural models provide adaptability but lack interpretability, while symbolic methods ensure explainability yet struggle with scalability in dynamic environments. This article examines the potential of Neuro-Symbolic Artificial Intelligence (NSAI) to bridge this gap by combining data-driven learning with logic-based reasoning. We outline the fundamentals of NSAI, explain its architectural integration into UAV communication systems across edge and cloud layers, and discuss its capacity to support interpretable, constraint-compliant decision-making in safety-critical missions. Representative use cases illustrate how NSAI can strengthen reliability, adaptability, and transparency in UAV operations. Finally, key challenges and open research directions, such as scalability, adversarial robustness, and alignment with Sixth-Generation (6G) and Non-Terrestrial Networks (NTN), are highlighted, offering a guideline for advancing NSAI as a pathway toward intelligent, resilient, and explainable aerial networks.

Reliable spectrum sensing (SS) is fundamental to cognitive radio (CR) and integrated sensing-and-communications (ISAC) systems. We propose a fluid antenna system (FAS)-aided improved energy detector (IED) that harnesses FAS port-selection diversity with a single RF chain. For an L-port FAS operating over correlated channels, we derive closed-form expressions for the average detection probability and the area under the receiver operating characteristic (ROC) curve (AUC) over fading and non-fading channels, along with their Gauss-Laguerre quadrature approximated expressions. Monte Carlo simulations validate the derivations and show consistent gains over classical energy detector (ED), highlighting the diversity benefits of FAS and the robustness advantages of p-norm detection. The results show that the IED with p=5 achieves a 7.8-times improvement over the classical ED. Moreover, the results quantify design trade-offs involving antenna size/spacing, number of ports, sensing time, threshold, and p, establishing FAS-aided IED as a low-hardware-complexity solution for robust SS in CR/ISAC.

Integrating unmanned aerial vehicle (UAV) based aerial reconfigurable intelligent surfaces (RISs) holds significant promise for enhancing the performance of ground-based networks. This paper introduces a novel 3D deployment and partitioning mechanism of aerial RIS for uplink grant-free non-orthogonal multiple access (GF-NOMA). GF-NOMA allows UEs to access the resource block immediately without waiting for grant acquisition. Specifically, we consider max-sum throughput (MST), max-min fairness (MMF), and proportional-fairness rate (PFR) objectives, jointly optimizing aerial-RIS partitioning and deployment, where we propose an unsupervised learning framework (USL) for partitioning and a deep reinforcement learning (DRL)-driven optimal deployment mechanism. Extensive numerical results show that the learned schemes for MST and MMF closely track their respective theoretical benchmarks, within 1% for MST and 2% for MMF. All three USL-DRL schemes significantly outperform fixed deployment baselines: the MST-USL-DRL achieves a 15-30% sum-rate gain, the MMF-USL-DRL improves the minimum user rate by 6-8%, and the PFR-USL-DRL delivers a 10-15% utility gain. Additionally, the PFR-USL-DRL surpasses the equal-allocation scheme by approximately 25%.

Machine learning (ML) has emerged as an effective approach for optimizing circuit design and bringing a paradigm shift in the development of wireless power transfer (WPT) systems. Being the main building blocks of near-field WPT, the slotted ground plane (SGP) resonators with a high quality factor (Q) enhance power transfer efficiency. However, it is pertinent to note that the resonator size, slot shape, and location result in distinct Q outcomes. Therefore, this work delves into the use of ML for predicting Q of SGP resonators. It can be predicted through a deep learning approach, owing to its capacity to learn from the implicit associations between input and output data. Hence, a deep neural network (DNN) model was designed using 20006 data files generated by electromagnetic (EM) simulations. DNN demonstrated its effectiveness, achieving an accuracy of 99.26%, thereby outperforming other benchmark ML models. Furthermore, the model proved its robustness in predicting Q of variously sized resonators and showed 98.3% accuracy. Subsequently, this enables anticipating the Q metric of scaled resonators without the need for exhaustive EM simulations. The predicted Q values were supported through experimental measurements. Finally, the SGP resonators were aptly employed to exhibit the near-field WPT system.