Channel knowledge map (CKM) is a key enabler for environmental awareness in future wireless systems. However, reconstructing high-fidelity CKM from sparse and noisy measurements poses a significant challenge. While Graph Neural Networks (GNNs) have emerged as a potent tool for this task, existing methods often lack physical consistency and generalization due to reliance on single graph structures and purely data-driven approaches. To tackle this challenge, in this paper, we propose a physics-informed multi-view graph neural network (PIMV-GNN) framework. This framework innovatively integrates two mechanisms within a unified GNN backbone: a multi-view learning (MVL) module that builds a rich spatial representation of the complex environment by fusing two complementary graph structures, namely a "spatial line-of-sight" graph and a "physical proximity" graph; and a physics-informed neural network (PINN) module that enforces physical consistency by imposing constraints derived from the Helmholtz equation in the graph domain. Extensive simulations demonstrate that our proposed PIMV-GNN framework significantly outperforms baseline models under various levels of data sparsity and measurement noise. Furthermore, the results reveal a profound synergistic effect between the MVL and PINN modules, where high-quality multi-view features significantly improve the regularization efficiency of the physics-based constraints.

NOMA-aided cognitive satellite aerial terrestrial integrated networks (CSATINs) are considered revolutionary and key technologies for 6G Internet of Things (6G-IoT), offering enhanced connectivity, high spectral efficiency, and broad coverage. In this article, we first establish trustworthy CSATINs with multiple aerial relays, aiming to achieve reliable communication in the presence of an eavesdropper. Then, to enhance the system’s covert performance, we propose an unmanned aerial vehicle scheduling scheme. Moreover, based on the established covert system model, we derive the closed-form expressions of detection error probability (DEP), covert outage probability (COP), and effective covert rate (ECR). Particularly, an optimization is proposed to enhance the covert performance of the considered system. Finally, Monte Carlo simulations are given to validate the correctness of the theoretical analysis, demonstrating that the reliability and covertness of the proposed system can be simultaneously enhanced by appropriately adjusting the power allocation coefficients, jamming power, and the transmission power of the satellite and UAVs.

This works investigates the covert performance for the satellite aerial ground integrated networks with imperfect limitations, i.e., channel estimation errors, non-ideal hardware and co-channel interference. To enhance the covert transmission, an unmanned aerial vehicle is applied to forward the signal from the satellite source to the destination. Particularly, the detection error probability, outage probability and the covert transmission rate are further studied in the presence of closed-form expressions and asymptotic expressions. Finally, some representative Monte Carlo simulations are proposed to verify our theoretical analysis. Derived from the results, the non-ideal hardware, the channel estimation errors and the co-channel interference have great impacts on the covert performance. Particularly, when the system is under non-ideal hardware, a suitable power allocation scheme is needed for the interference to gain the lowest detection error probability. Moreover, the outage probability has a lower bound in the case of non-ideal hardware. What’s more, the covert transmission rate also has an upper bound when the system is under non-ideal hardware. In addition, to have a better covert performance, the channel state information should be accurate enough by allocating much power for the channel estimations. The finding results are essential for the practical system for they can promote the engineering design.

Multi-reconfigurable intelligent surfaces (RIS) has recently gained significant interest as emerging technology for exploiting Intelligent electromagnetic environment. Inspired by opportunistic communications, a low-complexity adaptive selection and training method for multi-RISs is proposed in this paper. Firstly appropriate number of the multi-RISs is selected to assist communication. The elements of the selected RIS are grouped, and the grouped elements share a common coefficient to reduce training overhead. Secondly, the communication performance is evaluated and other RIS will be selected to assist communication if the communication performance not meet user requirement. In this way, the system performance and the training complexity of multi-RISs can be trade-off efficiently. Simulation results show that the proposed scheme outperforms the state-of-the-art benchmarks in terms of training overhead and robustness in different channel condition, and the training overhead is 30% lower compared to the centralized deployment scheme proposed in [13].

Covert communication has become a hot topic in the wireless transmission field due to the ability to secure transmitted data by hiding the wireless transmissions. Given the extensive use of drone communications and its urgent demand for security, we investigate covert communications in aerial terrestrial integrated networks (ATINs), where an unmanned aerial vehicle (UAV) tries to send private messages to a remote user via multiple terrestrial relays under the supervisions of the warden. On this foundation, one covert scheme for joint power control and relay selection has been proposed. Subsequently, we derive the detection capabilities at warden, and the effective covert rate (ECR) of link from UAV to user. Furthermore, a power optimization problem is designed to maximize ECR with covertness constraint. Finally, numerical results are presented to verify the achievable covert performance of system and prove the effectiveness of the proposed scheme.

ABSTRACT Different biometric modalities, such as fingerprints and left and right eye irises, contain physiological characteristics that offer high accuracy in identification processes. These modalities complement each other; for example, fingerprints provide intricate ridge patterns, while irises exhibit stable, precise features that perform well in challenging environments. A new proposed framework based on federated learning with optimised features, pre‐trained deep learning models, linear discriminant analysis and dense neural networks ensures privacy protection for multi‐modal biometric recognition across diverse biometric datasets. The system obtains better accuracy levels alongside increased robustness through the combination of fingerprint and iris scan technology that functions across independent and identically distributed (IID) and non‐independent and non‐identically distributed (non‐IID) conditions. Privacy protection functions as a key asset of federated learning because it allows distributed training operations through non‐raw data sharing, supporting high classification results. The system's performance is enhanced by implementing feature fusion alongside dimensionality reduction methods, which enhance both the efficiency and resistance to noise and variabilities. The system establishes an essential reference point for distributed and heterogeneous real‐world biometric recognition because it implements accurate computation with enhanced efficiency together with privacy protection. The IID data experiments demonstrated 98.86% training accuracy while achieving precision and recall at precise levels of 98.86% and 96.59%. All metrics achieved 100% on the validation data set while keeping loss at zero. The system's performance slightly decreased under non‐IID training data conditions, which resulted in 95.01% training accuracy and 0.18 training loss. The reported precision levels matched recall values since both measurements reached 97.99% and 95.01%. The system maintained perfect validation results through all metrics, which demonstrated a strong ability to generalise beyond data distribution impediments. The integration of multimodal biometric systems with federated learning enables the optimisation of large‐scale solutions because it establishes efficient but accurate and secure applications across domains that include surveillance and security together with healthcare.

Abstract Selecting crucial sentences from a document is a pivotal task in automatic text summarization systems. Abstractive summarization involves rephrasing key content through advanced natural language techniques, generating a concise, new text conveying critical information. Conversely, extractive summarization reproduces important material from the original text. In the proposed method, a hybrid ensemble approach combines BERTsum for extractive summarization and Longformer2Roberta for abstractive summarization for generating a contextual semantic rich summary for a huge collection of text. These proposed system‐generated summaries were evaluated against reference summaries using the ROUGE package at three rouge levels (Rouge‐1, Rouge‐2, and Rouge‐L). The Proposed contextual embedded hybrid text summarization model has shown significant performance improvement in multiple levels of Rouge score and word mover distance (WMD) of generated summary with a reference summary. The proposed hybrid model demonstrates superior performance over existing state‐of‐the‐art summarizing models on three distinct datasets CNN dataset, WikiSum, and Gigaword dataset. The proposed hybrid model as a text summarizer involves leveraging its capabilities to process longer sequences of text with domain‐specific contextual summaries. This transformers‐based text summarization model has great potential in developing expert systems in various research domains such as health decision support systems, the education sector, customer support chatbots, financial analysis investment recommendations, and financial assistance.

NR-V2X Mode 2 is introduced by the third generation partnership project (3GPP) to support vehicle-to-everything (V2X) communication. In NR-V2X Mode 2, vehicles select resources for the exchange of cooperative awareness messages (CAM) in a decentralized manner based on their local observation using semi-persistent scheduling. Resources are distributed over the 2-D frequency and time domain, following the long-term evolution frame structure. Since vehicles select resources based on their local observations and due to spectrum scarcity, this may lead to contention. Hence, selecting a resource is challenging, and as each vehicle strives to select a resource, it becomes a consensus problem. To resolve resource contention, in this article, we propose a knowledge-empowered distributed multiagent deep reinforcement learning (K-MADRL) approach. Based on traffic flow information, long short-term memory (LSTM) is employed to deploy Unmanned Internet of Aerial Agents (UIAAs) to collect vehicle state information. UIAAs gather vehicle state knowledge and train the local deep reinforcement learning (DRL) model. The locally trained model at the UIAA is shared and aggregated at the gNB for the global model update. The trained policy is then sent to the vehicles over system synchronization blocks for distributed execution. Moreover, the vehicles select the resource based on the joint action, i.e., by anticipating the actions of the neighboring vehicles. Our scheme is compared with other methods, such as DRL, optimization techniques, the SPS method, and random allocation methods, used in the NR-V2X environment. The results of the simulations show that our scheme outperforms the other methods.

This paper researches the secrecy outage probabili-ty for the rate splitting multiple access-based integrated satellite-aerial-terrestrial networks. Specially, owing to some practical reasons, imperfect hardware is further analyzed for all the network nodes. Moreover, a UAV is utilized to help the signal transmitting from the satellite to the ground destination in the presence of an eve. Besides, by considering these limitations, the detailed analysis for the secrecy outage probability are gotten, which offer a good way to calculate the impacts of channel parameters and system factors on the secrecy networks. Finally, several representative Monte Carlo simulations are presented to confirm the rightness of the analytical results.

The beginning of cellular communication (next-gen, such as 5G and 6G) promises an extreme leap in connectivity, introducing intelligent, adaptive solutions that integrate communication, artificial intelligence, and emerging technologies. Our approach combines reinforcement learning with Medium Access Control (MAC) protocols to dynamically optimize resource allocation and enhance network performance. In this work, we explore the integration of the adaptive frame size adjusting approach similar to the IEEE 802.1CB to ensure the efficient handling of seamless redundancy. The proposed solutions are validated through simulation, ensuring robustness and real-world applicability. Results indicate significant improvements in redundancy rate detection and delay in the network. This work contributes to achieving intelligent, adaptive, and seamless connectivity in the next generation of communication systems.

An exponential increase in the volume of connected user proximity wireless devices (UPWDs) is spearheading a hyper-connected ecosystem, which may enable smart cities, industries, and connected healthcare. However, this increase in the number of connected UPWDs results in significant amplification in electromagnetic field (EMF) exposure among users and consequently may result in potential physiological effects. Integrated sensing and communication (ISAC)-enabled ambient backscatter communication (ABC) is a promising technology that can power low-energy sensors and facilitate communication between data sources and sinks by reusing the available resources. The power-domain non-orthogonal multiple access (PD-NOMA) has the potential to provide channel resources to an increasing number of users simultaneously while adhering to the quality of service (QoS) requirements. However, empowering PD-NOMA with machine learning (ML) can mitigate challenges in massive channel access. This work uses a k-medoid and Silhouette analysis for sub-carrier allocation and optimizes power allocation to the users with optimization techniques using PD-NOMA in an ABC-enabled cellular network. The proposed system demonstrates a significant capability to reduce the aggregated uplink EMF exposure using robust and low-complexity ML techniques. The simulations show a superior performance compared to the state-of-the-art methods.

Reconfigurable intelligent surfaces (RISs) play a crucial role in improving the coverage and efficiency of millimeter-wave (mmWave) communication systems for Internet of Things (IoT) networks by enhancing signal strength and reducing interference. However, to fully exploit their potential, mathematical models and optimization algorithms are needed to optimize the RIS configuration and control. In this work, we present a mathematical model and optimization algorithm for RIS-aided mmWave communication systems. This paper proposes a mathematical model to capture the effects of RISs on mmWave communication channels, including equations for channel estimation and prediction. We also develop an opti-mization problem to maximize the system throughput, along with a model for optimizing the RIS phase shift using the alternating projection phase shift algorithm. The power allocation algorithm for the mmWave base station is presented using the Karush-Kuhn-Tucker (KKT) approach. Finally, we perform simulations to validate the proposed solution and compare it with ideal and random phase shift solutions. The results show that our proposed solution performs close to the ideal solution, demonstrating the effectiveness of the proposed mathematical model and optimization algorithm.

Smart cities intelligently control the functionalities of the city through the use of various electronic methods, sensors, and advanced communication techniques. An intelligent transport system (ITS) is the backbone of smart cities and refers to a system in which numerous vehicles utilize a communication infrastructure to exchange vital information, such as traffic, congestion, and road conditions. One of the key elements of ITS is vehicle tracking or localization, which is the monitoring of a moving vehicle’s location using a Global Positioning System (GPS). Accurate vehicle localization is required because many safety and traffic management applications depend on the precise positioning of the vehicles. Intelligent reflecting surface (IRS) is a new technology that offers attractive features like improved performance gains, enhanced network coverage, and flexible deployment, and is considered a key enabler for the 6G-driven vehicle-to-everything (V2X) systems that provide a programmable wireless environment. In this paper, a comprehensive view of a 6G-driven vehicle localization mechanism utilizing IRS is provided. The work also lists the benefits of IRS-enabled sensing in 6G vehicular networks including enhanced security, overcoming blockages, and improved localization. A case study to highlight the advantages of the IRS in vehicle tracking is presented. Finally, we also present the research challenges and future work directions in IRS-enabled vehicle tracking.

Abstract It is anticipated that sixth‐generation (6G) systems would present new security challenges while offering improved features and new directions for security in vehicular communication, which may result in the emergence of a new breed of adaptive and context‐aware security protocol. Physical layer security solutions can compete for low‐complexity, low‐delay, low‐footprint, adaptable, extensible, and context‐aware security schemes by leveraging the physical layer and introducing security controls. A novel physical layer security scheme that employs the concept of radio frequency fingerprinting (RF‐FP) for location estimation is proposed, wherein the RF‐FP values are collected at different points with in the cell. Then, based on the estimated location, the nearest possible road‐side unit for sending the information signal is located. After this, the effects on secrecy capacity (SC) and secrecy outage probability (SOP) in the presence of multiple eavesdropper per unit time are analysed. It has been shown via simulations that the proposed RF‐FP scheme increases SC by up to 25% for the same signal‐to‐noise ratio (SNR) values as those of the benchmarks, while the SOP tends to decrease by up to 30% as compared to the benchmark scheme for the same SNR value. Thus, the proposed RF‐FP‐based location estimation provides much better results as compared to the existing physical layer security schemes.

Lung cancer has been the leading cause of cancer death for many decades. With the advent of artificial intelligence, various machine learning models have been proposed for lung cancer detection (LCD). Typically, challenges in building an accurate LCD model are the small‐scale datasets, the poor generalizability to detect unseen data, and the selection of useful source domains and prioritization of multiple source domains for transfer learning. In this paper, a multiround transfer learning and modified generative adversarial network (MTL‐MGAN) algorithm is proposed for LCD. The MTL transfers the knowledge between the prioritized source domains and target domain to get rid of exhaust search of datasets prioritization among multiple datasets, maximizing the transferability with a multiround transfer learning process, and avoiding negative transfer via customization of loss functions in the aspects of domain, instance, and feature. In regard to the MGAN, it not only generates additional training data but also creates intermediate domains to bridge the gap between the source domains and target domains. 10 benchmark datasets are chosen for the performance evaluation and analysis of the MTL‐MGAN. The proposed algorithm has significantly improved the accuracy compared with related works. To examine the contributions of the individual components of the MTL‐MGAN, ablation studies are conducted to confirm the effectiveness of the prioritization algorithm, the MTL, the negative transfer avoidance via loss functions, and the MGAN. The research implications are to confirm the feasibility of multiround transfer learning to enhance the optimal solution of the target model and to provide a generic approach to bridge the gap between the source domain and target domain using MGAN.

To enable the internet of things (IoT) devices with increased coverage and optimized power consumption, the 3rdgeneration partnership (3GPP) standardizes the idea of narrowband IoT (NB-IoT) technology in the fifth generation (5G) of cellular communication. Re-transmission of control and data packets due to a poor link between user equipment (UE) and the base station (BS), is considered as one of the key feature in NB-IoT to ensure the data delivery of delay-sensitive applications, e.g. ambulance services and body sensor networks (BSN). This phenomenon degrades the energy efficiency of the already resource constrained systems. One key solution for NB-IoT UE is to exploit the device-to-device (D2D) communication using two hops instead of transmitting on a direct uplink, due to which the system performance increases. In an attempt to transmit the NB-IoT UE uplink data packet, splendid researchers have focused towards developing a D2D communication based strategy, which typically optimizes the expected packet delivery ratio (EDR) and end-to-end delay (EED) through an opportunistic method. However, such methodology imposes an additional delay due to the unavailability of active relaying nodes and increases the overall energy consumption of the system. This necessitates us to design an intelligent deterministic D2D (2D2D) relay selection strategy for delay sensitive NB-IoT UEs. The EDR and EED have been improved using deterministic programming based algorithm. Simulations with various parameters are carried out, and results are presented. Simulation results show that the deterministic algorithm gives better performance with a 10% increase in EDR and overcomes the additional delay.