Mazen Omar Hasna

Thermal noise modulation (TNM) encodes information in the variance of Johnson noise, enabling ultra-low/zeropower operation with intrinsic physical-layer security. While recent demonstrations have focused on binary TNM, its twolevel alphabet fundamentally limits spectral efficiency. This paper develops an M-ary TNM (M-TNM) framework that maps symbols to multiple variance levels, thereby conveying (log₂M) bits per channel use. We first establish a structural property of the optimal detector: for zero-mean Gaussian classes with ordered variances, the maximum-likelihood (ML) decision regions are contiguous in the energy statistic and each symbol’s boundaries depend only on its two adjacent variance levels. Leveraging this result, we derive closed-form expressions for the optimal ML detection thresholds and the associated exact average symbol error rate (SER) for M-TNM. We then formulate variance-constellation optimization—the selection and spacing of symbol variances—as a SER-minimization problem under practical constraints. Analytical results are validated via extensive Monte Carlo simulations, which show near-perfect agreement with theory and demonstrate substantial BES reductions for optimized constellations compared with naïve (e.g., uniformly spaced) variance levels. The proposed M-TNM framework thus improves spectral efficiency while preserving TNM’s energy and security promise, positioning variance-domain modulation as a viable physical-layer technique for bandwidth-limited, large-scale Internet of Things (IoT) deployments.

This paper presents an efficient analytical framework for evaluating the performance of inter-satellite communication systems utilizing orbital angular momentum (OAM) beams under pointing errors. An accurate analytical model is first developed to characterize intermodal crosstalk caused by beam misalignment in OAM-based inter-satellite links. Building upon this model, we derive efficient expressions to analyze and optimize system performance in terms of bit error rate (BER). Unlike traditional Monte Carlo-based methods that are computationally intensive, the proposed approach offers accurate performance predictions. This enables a substantial decrease in computation time while maintaining high accuracy, thanks to the use of analytical expressions for both crosstalk and BER. This fast and accurate evaluation capability is particularly critical for dynamic low Earth orbit (LEO) satellite constellations, where network topology and channel conditions change rapidly, requiring real-time link adaptation. Furthermore, we systematically design and evaluate asymmetric OAM mode sets, which significantly outperform symmetric configurations in the presence of pointing errors. Our results also reveal key insights into the interaction between beam divergence, tracking accuracy, and link distance, demonstrating that the proposed framework enables real-time optimization of system parameters with high fidelity. The analytical findings are rigorously validated against extensive Monte Carlo simulations, confirming their practical applicability for high-mobility optical wireless systems such as LEO satellite networks.

This paper addresses the joint problem of distributed UAV placement and multi-antenna beamforming for downlink Rate-Splitting Multiple Access (RSMA) in multi-UAV networks with uncertain user locations. We propose a practical two-stage hybrid solution that (i) learns robust, decentralized UAV placements via a QMIX-based multi-agent deep Q-learning framework that encodes the environment as fixed-resolution multi-channel spatial maps and employs CoordConv to preserve absolute spatial information, and (ii) optimizes per-UAV multi-antenna precoding via a tractable semidefinite-programming (SDP) relaxation augmented with successive convex approximation and difference-of-convex programming to handle nonconvex log-terms and rank-1 constraints. The learning stage provides centralized training with fully decentralized execution and yields a single trained agent that generalizes across varying user counts and cluster configurations without retraining. The convex precoding stage enforces user rate, power and RSMA SIC constraints and is solved efficiently using an interior-point SDP solver. Extensive simulations (varying user counts, UAV counts, grid resolutions and cluster patterns) demonstrate fast convergence of the proposed QMIX placement (convergence in 4×10⁴ training rounds in our setup), substantial sum rate gains over two competitive benchmarks, and favorable scalability and computational profiles.

Directional links in free-space optical (FSO), millimeter-wave (mmWave), and terahertz (THz) systems are a cornerstone of emerging 6G networks, yet their reliability is fundamentally limited by pointing errors and misalignment. Existing studies address this impairment using technology-specific definitions, models, and mitigation approaches, which hinders cross-domain comparison and transferable design insight. This survey provides a unified treatment of pointing errors across optical and high-frequency wireless communications. We establish consistent terminology and a cross-technology taxonomy of pointing errors, review angular misalignment and statistical distribution models, and analyze their impact on system performance. Mitigation techniques are systematically surveyed with emphasis on optical systems and their connection to underlying pointing-error models. The survey further provides a detailed examination of pointing-error effects in orbital angular momentum (OAM) links and quantum optical communications, and surveys the corresponding mitigation approaches tailored to mode-dependent impairments and quantum measurement constraints. The survey also outlines open challenges and future research directions. By consolidating fragmented literature into a coherent framework, this work supports consistent analysis and robust design of next-generation directional communication systems.

Unmanned Aerial Vehicles (UAVs) have emerged as a critical component in next-generation wireless networks, particularly for disaster recovery scenarios, due to their flexibility, mobility, and rapid deployment capabilities. This paper focuses on optimizing rotary-wing UAV trajectories to ensure effective communication in disaster-stricken areas using terahertz (THz) links. We address specific challenges such as energy consumption, user priority levels, and navigating complex urban environments to maintain Line of Sight (LoS) connections amidst 3D obstacles. Our contributions include the development of a detailed modeling approach using online 3D map data, the formulation of an optimal trajectory optimization problem, and the proposal of a Genetic Algorithm (GA)-based method alongside an enhanced heuristic algorithm for faster convergence. Through 3D simulations, we demonstrate the trade-off between minimizing total service time and prioritizing higher-weight nodes, showing the impact of different priority weight factors on the trajectory time. The proposed algorithms are evaluated using real-world data from the West Bay area of Doha, Qatar, demonstrating their effectiveness in optimizing UAV trajectories for emergency response.

This article investigates a movable-antenna (MA)-assisted dual-hop space–air–ground nonterrestrial network (NTN) operating under an underlay spectrum sharing paradigm. A satellite communicates with multiple uncrewed aerial vehicles (UAVs) over free-space optical (FSO) links, while each UAV simultaneously serves a cluster of ground users over radio frequency (RF) channels subject to interference constraints imposed by a primary receiver. To efficiently exploit the complementary advantages of FSO and RF transmission and the additional spatial degrees of freedom offered by MAs, a joint optimization framework is developed to maximize the system sum rate. The proposed framework jointly optimizes satellite and UAV power allocation, multiantenna precoding at the UAVs, and the positions of multiple MAs mounted on each UAV. An alternating-optimization algorithm is employed, where minimum mean square error (MMSE)-based precoding accounts for both communication and interference channels, power allocation is convexified using auxiliary rate variables and successive convex approximation (SCA), and antenna locations are optimized via Taylor-based convex surrogates. Simulation results demonstrate that the proposed approach significantly outperforms benchmark schemes with fixed antenna locations, heuristic optimization, and reinforcement learning methods, while providing robust performance across different system parameters.

In this article, we investigate the performance of wireless-powered communication (WPC) systems operating under the harvest-then-transmit protocol with a three-piecewise nonlinear energy harvesting (EH) model, which explicitly captures the sensitivity and saturation behavior of practical EH circuits. In parallel, we adopt a versatile fading characterization by assuming Nakagami-m channels in the EH phase and generalized Gamma channels in the information transfer phase, the latter being of practical interest as it unifies several standard fading distributions and accurately approximates fading environments in next-generation networks. Based on this setting, we develop a novel analytical framework that yields, for the first time, exact analytical expressions for the average throughput in the delay-limited, delay-tolerant, and quality-of-service (QoS) delay-constrained transmission modes. We then apply the developed framework to study the throughput of WPC-enabled reconfigurable intelligent surface-assisted systems as a direct application example. We conduct extensive Monte Carlo simulations under various system configurations to confirm the accuracy of the analytical results. Our results show that nonlinear EH effects yield a more realistic benchmark for the average throughput compared with the conventional linear EH model, particularly under stringent sensitivity and saturation constraints.

Reconfigurable intelligent surface (RIS)-based re f lection modulation (RM) is recognized as an innovative concept for effectively transferring information with more spectral effi ciency. This paper investigates the performance enhancements enabled by the quadrature reflection modulation (QRM) into uplink multiple access (MA) transmission. Specifically, we pro pose a novel QRM indexing-aided RIS grouping design for the MA system, termed QRMMA, to boost the system's spectral efficiency and enhance the error performance. The main idea is to separate the RIS groups into two orthogonal subsets equally, then each subset is further divided into pairs, where each user should select a single subgroup from each pair based on its transmitted data. To evaluate the performance of the proposed system, we derive an approximate expression for the average bit error probability (ABEP) and ergodic capacity over Rician fading channels. Simulation results demonstrate that the proposed QRMMA system can achieve a significantly enhanced performance compared to its conventional counterparts while maintaining higher spectral efficiency.

Non-terrestrial networks (NTNs) leveraging high-altitude platforms (HAPs) are pivotal for bridging coverage gaps in remote regions and ensuring service continuity when terrestrial infrastructure fails (e.g., during disasters) or becomes congested (e.g., mass events). While integrating rate-splitting multiple access (RSMA) and reconfigurable intelligent surfaces (RISs) enhances spectral efficiency, securing transmissions against unknown eavesdroppers with imperfect channel state information (CSI) remains a critical challenge. We propose a secure multi-user, multi-cluster RIS-assisted RSMA system for HAPs, jointly optimizing precoding, RIS beamforming, and cluster time allocation to maximize sum-rate under secrecy rate constraints. The non-convex problem is transformed into a tractable semidefinite program (SDP) via successive convex approximation, auxiliary variable bounding, and a rank-1 penalty method. Simulations demonstrate significant performance gains over benchmarks, robustness to practical RIS phase quantization, and scalability across varying numbers of users per cluster, validating the framework's efficacy.