Covert communication has sparked widespread attention since it seeks to secure potentially undetected transmissions by employing a variety of confusion sources. Despite the highly directed nature of Terahertz (THz) beams, receiver vibration causes beam direction misalignment, lowering communication speeds, and reducing communication covertness. In this research, we present and analyze a resilient reconfigurable intelligent surface (RIS)-aided covert transmission scheme for THz transmissions, in which a multi-antenna node transmits to both a covert user and a public user in the presence of an unfriendly warden, under both perfect and imperfect channel state information assumptions. The proposed approach is based on superimposing the two legitimate users’ signals, confusing the warden node’s detection. Closed-form expressions for the optimal detection threshold at the warden, detection error probability (DEP), and minimum DEP (MDEP) are provided. Simulation findings corroborate the analytical results, and insightful discussions on the monotonicity of the main metrics and their impact on the system’s covertness are included. Then, using an alternating optimization (AO)-based framework, an optimal design of the power allocated to each user’s signal, their respective beamforming vectors, and the RIS phase shifts is carried out to maximize the system’s achievable communication rate subject to power constraints and an MDEP threshold at the warden. The convexity of all intermediate subproblems has been thoroughly examined, and the covertness advantage of the proposed approach has been proven.

This paper investigates covert communication in reconfigurable intelligent surface (RIS)-assisted wireless systems when an adversarial detector (Willie) patrols the environment along a circular trajectory. Unlike prior works assuming stationary eavesdroppers, we explicitly model adversary mobility and derive analytical feasibility conditions for covert transmission under average leakage constraints. Leveraging Rayleigh quotient optimization with unimodular constraints, we obtain a closed-form RIS phase design maximizing legitimate received power while satisfying covert constraints. Extensive Monte Carlo simulations validate the theoretical predictions, revealing a fundamental trade-off between Willie’s mobility radius and achievable performance at Bob. Sensitivity analyses demonstrate that RIS size and covert constraint tightness critically impact system behavior, confirming the importance of modeling adversary mobility in practical RIS-assisted covert communication systems.

Federated Learning in Internet of Things (IoT) networks involves a critical trade-off between resource efficiency and global model performance. We propose a lightweight optimal stopping theory-assisted node selection strategy, which unifies local model accuracy (LMA) and received signal strength (RSS) into a single scoring metric governed by a tunable parameter α. The optimized value α maximizing the probability of picking the best elements following our strategy has been obtained through a probabilistic framework. The analytical results have been confirmed through simulation. Experiments have been conducted employing the MNIST dataset and show consistent improvements in global model accuracy compared to four baseline strategies, while ensuring a suitable complexity for real-time IoT deployment. Moreover, unlike clustering-based or secretary problem-inspired approaches, our method evaluates all nodes globally and performs reliably even with limited candidate availability.

Implementing Federated Learning (FL) in IoT environments presents notable challenges, primarily due to variations in both clients' local model performance and the reliability of their communication links, often indicated by received signal strength (RSS). In this study, we introduce a deterministic and lightweight selection mechanism that filters participants based on predefined thresholds for local model accuracy (LMA) and RSS. The thresholds are analytically optimized to maximize the probability of selecting exactly$m$suitable clients. Theoretical analysis is complemented by parameter-specific insights to better illustrate the system's behavior. Comprehensive experiments using the MNIST dataset demonstrate that our approach consistently surpasses strategies that rely solely on LMA or RSS, validating its effectiveness for FL in resource-limited IoT contexts.