Ensuring secure and efficient surveillance in maritime border security is critical to addressing threats such as illegal trafficking, unauthorized vessel movements, and piracy. This paper presents a novel Autonomous Aerial Vehicle (AAV)-assisted surveillance framework that leverages QoS-aware edge computing to enhance real-time situational awareness, task offloading, and secure trajectory optimization. The proposed system integrates Twin-Delayed Deep Deterministic Policy Gradient (TD3) reinforcement learning for adaptive AAV trajectory planning, ensuring optimal coverage and minimal energy consumption. Enhanced Particle Swarm Optimization (EPSO) is also employed for intelligent task offloading, efficiently balancing computational workloads between AAVs and edge nodes. It is evaluated through simulations with real-world maritime surveillance scenarios, demonstrating reduced latency and improved energy efficiency compared to conventional surveillance and task management strategies.

Green communication and sustainable operations have become critical objectives in Information and Communications Technology (ICT) systems, particularly when integrating energy-intensive technologies such as Autonomous Aerial Vehicles (AAVs). Therefore, this paper introduces a dynamic resource allocation framework for AAV-enabled green edge networks that adaptively manages bandwidth and computational power while optimizing AAV trajectories. By explicitly formulating the problem as a Markov Decision Process (MDP) and employing Deep Reinforcement Learning (DRL) with Proximal Policy Optimization (PPO), the proposed system strikes a balance between high data synchronization demands and strict energy constraints, leading to improved throughput and sustainability. The simulation results reveal that this approach significantly boosts data throughput and communication efficiency while reducing energy consumption. These findings pave the way for environmentally responsible edge networks that meet both performance requirements and sustainability targets.

The proposed system known as Vision Transformer-Masked Autoencoder-Cheetah Optimization Algorithm (ViT-MAE-COA) uses image preprocessing techniques in addition to segmentation and classification features and hyperparameter optimization capabilities to classify Otitis Media. The framework starts by improving image quality through the Wiener filter (WF) that minimizes mean squared error between original images and noisy images for noise reduction purposes. The W-Net architecture processes segmented data to maintain essential localization data and content information through a strategy that decreases parameters with max pooling. The results indicate that the model exhibited better prediction accuracy than other Deep Learning models.

The Internet of Medical Things (IoMT) is transforming healthcare by enabling real-time monitoring, diagnostics, and secure data-driven decision-making. However, IoMT networks are vulnerable to adversarial attacks, data breaches, and privacy threats, making secure and optimized node deployment a critical challenge. This paper presents a novel framework integrating 3D Voronoi diagrams and K-means clustering with a hybrid Particle Swarm Optimization-Genetic Algorithm (PSOGA) to optimize IoMT node placement while enhancing security and resilience. Initially, K-means clustering distributes nodes, followed by spatial partitioning with 3D Voronoi diagrams. The PSO-GA hybrid algorithm then iteratively refines node positions, balancing rapid convergence with global exploration to achieve optimal configurations that improve coverage, energy efficiency, and secure data exchange. Additionally, the proposed approach integrates risk assessment techniques and privacypreserving mechanisms to mitigate adversarial threats, ensuring robustness against poisoning and evasion attacks. By dynamically adapting to changing healthcare environments, the framework enhances network resiliency while aligning with AI security and privacy-by-design principles. Experimental results validate the algorithm's scalability and effectiveness, making it a promising solution for real-world IoMT applications in secure medical monitoring, diagnostics, and AI-driven threat intelligence.

In Intelligent Transportation Systems (ITS), driver recognition presents challenges of data privacy, computational efficiency, and energy consumption. Optimizing energy use in ITS has become crucial with the rise of environmentally conscious Information and Communication Technology (ICT) practices. Therefore, this paper introduces a privacy-preserving and energy-efficient task offloading strategy using Deep Reinforcement Learning (DRL) and Federated Learning (FL) within a network leveraging Smart Traffic Cameras (STCs) for edge computing. Initially, the public transports employ a DRL-based strategy to offload tasks to STCs, optimizing network resources and minimizing energy use. At the second phase, allows private vehicles to train models locally, offloading only model parameters, thus ensuring data privacy and reducing communication energy costs. Finally, aggregates these parameters at a central cloud server, refining a Network-Wide Model (NWM). The proposed framework enhances model performance, preserves privacy, and improves computational efficiency, reducing the carbon footprint of ITS operations. Simulations demonstrate that DRL's Actor-Critic Algorithm (ACA) reduces task latency and energy consumption while FL ensures efficient model training with minimal communication overhead.

Network slicing has emerged as a transformative enabler for meeting the diverse requirements of 5G and beyond networks, including 6G. However, network slices’ dynamic and virtualized nature introduces significant security challenges, particularly against evolving cyber threats. We propose a Deep Reinforcement Learning (DRL)–based Moving Target Defense (MTD) strategy tailored for secure network slicing to address these challenges. Our approach utilizes a Q-Learning framework to manage MTD actions dynamically, optimizing security while maintaining service quality. Extensive simulations demonstrate the effectiveness of our framework in minimizing attack success rates and ensuring operational stability, significantly outperforming baseline methods such as random decision-making.

Modern IoT and Fog environments are complex and diverse ecosystems that consist of numerous devices. Some of these devices can receive and process offloaded tasks. For such devices to operate at the highest capacity levels, there is a need for mechanisms that could optimize their performance with offloaded tasks. That includes, but is not limited to, such aspects as resource management, workload balancing and scheduling. Unlike local tasks, offloaded ones are not a part of device’s environment. Therefore, processing them should not irreparably disrupt a device’s functionality. This requires devices to have a mechanism for managing offloaded tasks differently from their local. The current work attempts to research possible ways to optimize in-device execution of offloaded tasks, while reducing detrimental effects to a device’s state. To achieve that, the solution involves application of Reinforcement Learning techniques. The work proposes to utilize Deep Deterministic Policy Gradient (DDPG) Actor/Critic method, to allow devices continuously learn optimal scheduling strategies for offloaded tasks. The contribution of this work is in its exploration of the impact machine learning makes on in-device scheduling, application feasibility and the overall execution time optimization of offloaded tasks.

Smart spaces integrate advanced technologies like the Internet of Things (IoT), Machine Learning, and Artificial Intelligence (AI) to enhance automation and control within various environments. Effective deployment of IoT nodes is crucial for maximizing coverage, minimizing costs, and ensuring network stability in these spaces. This paper presents a novel approach combining 3D Voronoi diagrams with a modified Genetic Algorithm (GA) to optimize IoT node placement in three-dimensional environments. The proposed method starts with node placement using a homogeneous Poisson Point Process (PPP) and partitions the space into Voronoi cells, followed by iterative adjustments using the modified GA. The method achieves a 15% improvement in coverage ratio, a 10% reduction in deployment effort, and a 20% increase in network stability compared to existing algorithms, with results statistically significant at 5%. Moreover, optimising sensor placements indirectly enhances network security by reducing redundant data paths and strengthening network resilience. This study provides a scalable, efficient solution for IoT network deployment in complex environments, addressing key challenges in smart space optimization and paving the way for more secure and robust IoT infrastructures.

Integrating Unmanned Aerial Vehicles (UAVs) into the emerging sixth-generation and beyond (6G+) cellular networks as aerial base stations represents a significant technological advancement. This integration offers numerous benefits, including widespread accessibility, enhanced navigation, and simplified monitoring and management. A key element of this integration involves the instantaneous distribution of vital information throughout the transportation infrastructure. Characterized by their agility, mobility, and flexibility, UAVs play a crucial role in relieving data traffic loads, thereby offering additional access points. This function is essential for making prompt, precise, and well-informed decisions in Intelligent Transportation Systems (ITS), utilizing data-centric insights. Deploying versatile Road Side Units (RSUs) for secure data collection and dissemination requires a robust framework for safe data transfer. Ensuring data governance in the Internet of Vehicles (IoV) network relies heavily on specific interactions between trusted parties. In response, we introduce an advanced encryption approach to promote secure data exchange in ITS, thus supporting the confidential transfer of information in IoV communications. This innovative encryption method can also perform encryption and decryption of ciphertexts, encompassing confidential data and facilitating secure communication.

This study presents an innovative approach to enhancing the efficiency of Unmanned Aerial Vehicles (UAV) in IoT networks. Employing UAVs as flying relays focuses on their role in data collection and support for terrestrial cellular networks. The central innovation lies in the application of Federated Learning (FL), which processes data while ensuring user privacy and reducing communication overhead. Addressing the challenge of UAVs’ limited battery life, which restricts sustained FL operations, we introduce the Enhanced UAV Network optimization Algorithm with Adaptive Spatial Play (ENUO-ASP). ENUOASP incorporates a modified Particle Swarm optimization (PSO) technique to determine optimal UAV placements, enhancing data collection by focusing on the Signal-to-Interference Ratio (SINR). Additionally, the paper utilizes the Deep Deterministic Policy Gradient (DDPG) method for dynamic resource allocation, optimizing energy consumption and reducing link latency between the UAV network and users. The findings indicate that the ENUO algorithm outperforms existing methods by achieving higher data rates and balanced SINR. Furthermore, the ASP resource allocation strategy improves FL execution, significantly lowering latency and energy use. This research contributes to the UAV-enabled communication field, offering a more efficient and performance-driven solution for advanced IoT applications.

This paper explores the integration of Quantum Communication Networks (QCNs), semantic communication, and covert communication within the context of 6G and future wireless networks. Introducing a new Quantum Semantic Communications (QSC) framework that capitalizes on advancements in quantum machine learning and semantic representations, this framework dramatically enhances resource efficiency in QCN s. It does so by embedding only relevant classical data into compact, high-dimensional quantum states for transmission, achieving a potential resource reduction of 50-75% while boosting quantum semantic fidelity. The paper further examines Artificial Intelligence's (AI) transformative role in upgrading tra-ditional communication paradigms into more efficient semantic communication systems, utilizing deep learning and end-to-end methodologies to ensure precise conveyance and interpretation of meanings in transmitted information. Additionally, it explores incorporating covert communication strategies within systems supported by a Reconfigurable Intelligent Surface (STAR-RIS) and Non-Orthogonal Multiple Access (NOMA), emphasizing the enhanced security and stealth necessary for modern networks. By merging these sophisticated communication strategies, the paper anticipates a new era of telecommunications that significantly surpasses existing security, efficiency, and semantic accuracy capabilities, marking a progressive step towards future networks optimized for secure, efficient, and meaning-focused communication in the quantum and AI era.

The high proliferation of on-line gaming along with the high demands of availability of network resources, created the need for the development of Cloudified services that will augment computation capabilities of mobile devices. To this end, this work elaborates on the design, the development and the comparative evaluation with other similar models, as well as with real-time comparisons through emulators, of a process-offloading scheme that is based on a mobile opportunistic cloud computing approach. According to the proposed approach, each mobile device with access to interactive -delay sensitive- multimedia content (i.e. online gaming with processing power requirements) exploits several network-centric parameters, by using Nano-Mobile Data Centers for an interactive, collaborative and real-time manipulation of the available resources. The communication and the social context is used by the mobile nodes with other communication related parameters, towards achieving the efficient execution of the offloading process in order to support adequate quality of service. The proposed scheme allows interactive mobile users to efficiently exploit their resources, while the processes that cannot be locally handled (by each device), are effectively offloaded. The scheme aims at prolonging the lifetime of each mobile device and maximizing the efficiency in running context interactive applications. The efficiency of the proposed scheme is validated through comparative performance evaluations with other similar schemes, indicating the level of the mobile nodes lifetime extensibility that is offered, in contrast to existing approaches.

This paper proposes a delay-assisted cooperative scheme for optimal TV White Spaces (TVWS) exploitation and maximum energy conservation in a 5G mobile cognitive radio (CR) network architecture. This architecture utilizes a radio spectrum broker, which administrates the process of network resources management among several 5G base-stations to support Quality of Service (QoS) provision and minimum energy consumption. The proposed scheme is based on the comparison of the delays of both the secondary nodes and the Radio Access Points, when a delay sensitive transmission is requested. The validity of the proposed scheme is verified through several experimental tests, performed under controlled simulation conditions. The performance evaluation results include the energy consumption level and the lifetime span of each wireless node, the throughput response of the system during the delay-sensitive resource exchange process, as well as quantitative measurements of the energy efficiency levels of the proposed scheme.

This paper proposes two radio resource management (RRM) algorithms for efficient QoS provisioning over an infrastructure-based cognitive radio network architecture that enables for TV White Spaces exploitation. QoS provisioning and policy management is achieved via a spectrum broker that coordinates the RRM process among LTE secondary systems, under the real time secondary spectrum market policy. The proposed RRM algorithms administrate the economics of the transactions between the spectrum broker and secondary systems, following a fixed-price and an auction-based trading process. The validity of the proposed algorithms is verified via a number of tests, carried under controlled experimental conditions (i.e. simulations), evaluating spectrum broker benefit and secondary systems service rate.

This paper proposes an energy-efficient routing scheme that enables energy conservation and efficient data flow coordination, among communication nodes with heterogeneous spectrum availability in distributed cognitive radio networks. Effective routing scheme operation, as a matter of maximum energy conservation and traffic manipulation is obtained, by utilizing backward traffic activity evaluation, developed based on a simulation scenario. This simulation scenario includes a number of secondary communication nodes, operating over television white spaces (TVWS), under “spectrum of commons” regulation policy. The validity of the proposed energy-efficient routing scheme is verified, by conducting experimental simulations and obtaining performance evaluation results. Simulation results validated its efficiency for minimizing energy consumption and maximizing resources exchange among secondary communication nodes.

The paper presents an infrastructure-based cognitive radio network architecture that enables for TV white spaces exploitation, QoS provisioning and policy management, under the real time secondary spectrum market policy. It describes the configuration of a spectrum broker that coordinates the radio resource management process (RRM) among LTE secondary systems as a matter of maximum possible TVWS utilisation and minimum frequency fragmentation, and also administrates the economics of such transactions towards maximum revenue following a fixed-price trading. The validity of the proposed architecture is verified via a number of tests carried under controlled experimental conditions (i.e. simulations) exploiting a decision-making algorithm.

The paper discusses the TV white spaces exploitation by a prototype centralised cognitive radio network architecture, under the real time secondary spectrum management scheme. Vital part of this architecture is a spectrum broker that coordinates the radio resources allocation process among secondary systems, as well as the transactions of spectrum trading following a fixed-price policy. Efficient broker operation as a matter of maximum-possible spectrum utilisation and minimum fragmentation is obtained by decision-making methods based on Backtracking, Simulated Annealing and Genetic algorithm. The validity of the proposed approach is verified via a number of experiments under controlled conditions, while its performance is evaluated against a number of secondary systems competing for TVWS exploitation, each one featuring different transmission characteristics.

The paper proposes a routing protocol that efficiently coordinates data flows among secondary systems with heterogeneous spectrum availability in an ad-hoc cognitive radio network architecture. Efficient protocol operation as a matter of maximum-possible routing paths establishments and minimum delays is obtained by a coordination mechanism, which was implemented based on a simulation scenario. The simulation scenario includes a number of secondary systems that exploit television white spaces, under the spectrum of commons regime. The validity of the research approach is verified via a number of experimental tests, conducted under controlled simulation conditions, evaluating the performance of the proposed routing protocol.