Search Results (1,054)

Flying Ad Hoc Networks (FANETs) composed of multiple unmanned aerial vehicles (UAVs) are a promising solution for emergency wireless communications when terrestrial infrastructure is unavailable. However, ensuring reliable Quality of Service (QoS) in these highly dynamic networks remains challenging due to topology changes, varying propagation conditions, and congestion. This work proposes a lightweight machine learning-based QoS optimization framework for multi-UAV emergency communications that combines realistic mobility modeling, empirical channel measurements, and adaptive traffic prioritization. UAV mobility patterns are generated with ArduSim, while LoS/NLoS propagation models are derived from real UAV flight experiments and integrated into ns-3. Multiple supervised machine learning algorithms—including Decision Trees, Random Forest, Support Vector Machines, k-NN, Gradient Boosting, and CatBoost—are trained using four input features derived from the network state: ${\mathrm{CBR}}_{\mathrm{src}}$, ${\mathrm{QP}}_{\mathrm{src}}$, ${\mathrm{CBR}}_{\mathrm{dst}}$, and ${\mathrm{QP}}_{\mathrm{dst}}$. Simulation results show that the proposed AI SMOTE EMERGENCY scheme, based on CatBoost, improves the Packet Delivery Ratio (PDR) by approximately 43% over the No-QoS baseline, achieving 89–93% delivery across all four application ports. Compared with EDCA, the proposed scheme maintains reliable delivery for all services, increases emergency throughput by 34–36%, and reduces end-to-end delay by about 70%. In addition, the higher delivery reliability translates into clear communication energy benefits, reducing energy waste across all evaluated topologies when compared with the No-QoS baseline. The inference time remains below 0.002 s, supporting real-time QoS adaptation in resource-constrained UAV networks. Full article

To support intelligent maritime applications, space–air–ground–sea integrated networks (SAGSINs) have been introduced in maritime communications to provide wide coverage and reliable network services. In unmanned aerial vehicle (UAV)-assisted SAGSIN architectures, UAVs can flexibly extend coverage and provide on-demand communication and computing support. However, due to the high mobility of low Earth orbit (LEO) satellites and the limited endurance of UAVs, single-platform deployment strategies struggle to provide both flexibility and scalability in maritime communication networks. To mitigate the service instability caused by satellite orbital dynamics and limited UAV endurance, we propose a Hybrid Multi-Agent Proximal Policy Optimization (H-MAPPO)-based joint satellite–UAV deployment scheme for UAV-assisted SAGSIN systems. The proposed method optimizes joint UAV positioning and resource allocation to enhance communication coverage while reducing overall operational cost. By incorporating satellite orbital dynamics and UAV mobility into a multi-agent reinforcement learning (MARL) framework, adaptive resource scheduling can be achieved under time-varying maritime demands. Simulation results show that the proposed H-MAPPO algorithm achieves superior convergence performance, higher user coverage, and lower total system cost compared with learning-based, random, and heuristic methods while maintaining stable and robust performance under varying user densities and network scales. Full article

This study examined the implementation and instructional effectiveness of a multimodal AI-supported instructional framework integrating a generative AI assistant (Microsoft Copilot) with a speech-recognition-based mobile learning application (Mondly) to support productive vocabulary development in EFL higher education. Unlike studies focusing on single AI tools, this study evaluates a coordinated dual-module instructional configuration combining LLM-based lexical support with ASR-based spoken retrieval practice within a structured classroom routine. The proposed framework can be viewed as a lightweight socio-technical instructional arrangement in which learners engage with complementary AI components through guided feedback and repeated practice. A quasi-experimental pretest–post-test control group design was conducted over an eleven-week semester with 64 first-year EFL students at an Iraqi university. Productive vocabulary knowledge was measured using the Productive Vocabulary Levels Test (PVLT), and data were analyzed using mixed-design ANOVA. Results revealed a statistically significant Time × Group interaction with a large effect size, indicating greater productive vocabulary gains in the AI-supported condition compared with traditional instruction. Qualitative findings further suggested perceived improvements in lexical retrieval, sentence construction, pronunciation accuracy, and learner engagement. From an instructional perspective, the findings suggest that learning gains were associated with the coordinated use of complementary AI tools within a structured classroom workflow. This study provides a practical instructional model that may be adaptable to comparable resource-constrained higher-education contexts. Full article

Edge computing has emerged as a promising paradigm to minimize latency and energy consumption while improving computational efficiency for mobile devices. Latency-sensitive applications such as autonomous driving, augmented reality, and industrial automation require ultra-low response times, making efficient task offloading a necessity in edge computing. However, distributing optimally computational tasks among edge servers remains a challenge, especially when considering latency, energy consumption, and workload balancing simultaneously. Although existing approaches have focused on one or two of these objectives, they do not provide a holistic solution that incorporates all three factors. In addition, some existing solutions do not take advantage of parallelism at the edge layer, resulting in bottlenecks and inefficient resource usage. In this paper, we propose a novel learning-based task offloading model that integrates parallel processing at the edge layer, adaptive workload balancing, and joint latency–energy optimization. Moreover, by dynamically adjusting the number of selected edge servers for parallel execution, our approach achieves optimal trade-offs between performance and resource efficiency. Our experimental setup includes several edge servers and several randomly deployed devices. It employs Apache HTTP Benchmark (AB) to generate realistic Mobile Edge Computing workloads. The obtained results show that our method outperforms existing approaches by reducing latency, lowering energy consumption, and maintaining a balanced workload across edge nodes. Full article

The high accuracy achieved by deep learning techniques has made them indispensable in computer vision applications. However, their substantial memory demands and high computational complexity limit their deployment in resource-constrained environments. To address this challenge, this study introduces a Feature Enhancement Module (FEM) as part of a unified hybrid compression framework that combines mixed-precision quantization and structured pruning to improve model efficiency. Experimental results on the Tiny ImageNet dataset using ResNet50 and MobileNetV3 architectures demonstrate the strong adaptability and scalability of the proposed approach. Compared with state-of-the-art compression methods, the proposed FEM-based framework achieves up to 6% improvement in Top-1 accuracy, while reducing memory usage by 32.26% and improving inference speed by 66%. Furthermore, the ablation study demonstrates that incorporating the FEM module leads to up to 24% improvement over the baseline model, highlighting its effectiveness. The results further show that FEM effectively preserves inter-channel feature representation stability even under aggressive compression, making it well suited for real-time processing and practical Artificial Intelligence (AI) applications. By maintaining semantic richness while significantly reducing computational cost, the proposed method bridges the gap between high-performance deep models and lightweight, deployable solutions. Overall, the FEM-based hybrid compression framework establishes a scalable and architecture-independent foundation for sustainable deep learning in resource-limited environments. Full article

Background/Objectives: The American Heart Association Life’s Essential 8 (LE8) is a tool proposed to categorize overall cardiovascular health (CVH), ranging from 0 to 100 and classifies CVH as low (<50), moderate (50–79) or high (≥80), based on the following health behaviors (diet, physical activity, nicotine exposure and sleep) and health factors (body mass index—BMI, lipid levels, glycemic profile and blood pressure). Although used in the general population, it is not part of the health assessment routine in the workplace. We assessed CVH of healthcare workers using an LE8-based score through a mobile application. Methods: Cross-sectional pilot study with adults working at a tertiary hospital in Brazil. We used an app for self-reporting LE8 metrics. Additionally, data on age, sex, and mental health (10-item Perceived Stress Scale, PSS-10) were collected. Results: Sixty-five adults (58.5% female; mean age 36 ± 9.01 years) were included. The mean LE8 overall score was 69.39 ± 12.63. The proportion of participants in the low, moderate and high cardiovascular health categories were 6.2%, 69.2% and 24.6%, respectively. Diet quality (34.76 ± 24.3) and physical activity (45.38 ± 40.58) were in the “low cardiovascular health” category. “Health behaviors” had a significantly lower mean score than “health factors” (58.90 ± 20.53 vs. 79.88 ± 15.55, p < 0.001). The mean PSS-10 score was 19.01 ± 7.49, indicating moderate perceived stress. Overall LE8 and PSS-10 scores were not significantly correlated (r_s = −0,0.17; p = 0.161). There was no significant difference in the mean overall LE8 score in the linear regression model adjusting for age, sex and perceived stress. Conclusions: Among employees of a Brazilian tertiary hospital, the adapted LE8 score indicated overall moderate CVH. Health behaviors—particularly diet quality and physical activity—were the main vulnerable areas. Implementation of an LE8-based assessment in the workplace may be useful for targeted prevention strategies in Brazil. Future larger and longitudinal studies are warranted to confirm these findings. Full article

While direct current (DC) ice-melting is currently adopted for some transmission lines, its application to 10 kV distribution transformers—often located in remote and rugged terrain—presents significant operational challenges. Disconnecting these transformers prior to ice-melting is a complex procedure that incurs substantial labor, material, and financial costs. Leaving transformers connected risks DC current flowing into idle windings, potentially causing damage. Furthermore, existing mobile DC ice-melting power supplies are bulky and impose stringent transportation requirements, rendering them unsuitable for use on mountain roads. To overcome these limitations, this paper proposes a compact, lightweight variable-frequency ice-melting device. The operating principle and output characteristics of the variable-frequency method are investigated in detail. Using Simulink, system modeling and simulation analyses are performed to obtain the voltage and current output characteristics, along with harmonic spectra. Simulation results demonstrate that the proposed device achieves significant miniaturization compared with conventional solutions: within the typical parameter range of conventional devices, the volume can be reduced by 44–58% and the weight by 43–52%. In addition, the selected LC filter parameters (L = 10.39 mH, C = 86.62 μF) represent an optimized compromise solution that effectively suppresses input harmonics while maintaining the output current total harmonic distortion (THD) within an acceptable limit of 3.6%. Experimental results further validate the feasibility of the variable-frequency ice-melting current. Based on a matrix converter topology, the proposed device enables flexible adjustment of the output melting voltage and frequency, exhibits excellent low-frequency performance and dynamic response, and maintains low output harmonic content—fully meeting the application requirements for variable-frequency ice-melting. The key novelty lies in a compact matrix-converter-based de-icing device with systematic low-frequency performance analysis, offering superior portability and adaptability over traditional DC solutions. Full article

Diameter at Breast Height (DBH) is a key parameter in forest measurement. However, existing research has mostly focused on improving the accuracy of individual technologies, lacking a systematic synthesis of the evolutionary logic of measurement techniques and a standardized selection framework for forestry applications. To this end, this paper constructs a multi-level classification framework based on measurement platforms and technical principles, establishes for the first time a five-dimensional comprehensive evaluation system (covering accuracy, efficiency, cost, environmental adaptability, and automation) along with a hierarchical technology decision tree, and systematically analyzes the application logic of multi-source fusion technologies across three levels: ground-based, near-ground mobile, and aerial. The review indicates that traditional contact-based measurement has limited efficiency; modern remote sensing technologies (photogrammetry and LiDAR) offer significant advantages in automation and accuracy, but still face challenges such as high equipment costs, complex data processing, and poor environmental adaptability. Multi-source fusion and machine learning are key methods to overcome the limitations of single sensors and improve the robustness of DBH estimation. Finally, it is anticipated that with decreasing sensor costs and the advancement of intelligent algorithms, DBH measurement will continue to evolve toward automation, intelligence, and engineering practicality, providing technical support for large-scale, long-term, and repeatable forest monitoring. Full article

To investigate the key role of endophytic fungi in maintaining host adaptability and overall health, endophytic fungi were isolated from healthy root, stem and leaf tissues of Impatiens hawkeri, and the dominant strain FG8 with growth-promoting and antagonistic functions was screened. Strain FG8 was identified as Thelonectria veuillotiana by morphological and molecular biological methods. It exhibited an antifungal rate of 58.57% against Stagonosporopsis cucurbitacearum, the pathogen causing leaf spot disease of I. hawkeri. The broad-spectrum antifungal activity was verified by the plate confrontation method, and FG8 showed inhibitory effects on six common pathogenic fungi, with the highest inhibition rate of 64.5% against Apiospora intestini. Furthermore, strain FG8 displayed remarkable growth-promoting and antagonistic characteristics: it produced indole-3-acetic acid at 12.74 μg/mL, and possessed the abilities of phosphate solubilization, potassium mobilization, nitrogen fixation and siderophore synthesis. Its antagonistic activity was mediated by β-glucanase, amylase, cellulase and pectinase. Meanwhile, FG8 significantly induced the activities of four defensive enzymes in I. hawkeri, including superoxide dismutase (SOD), peroxidase (POD), catalase (CAT) and polyphenol oxidase (PPO). Seed growth-promotion experiments demonstrated that the root length, plant height, fresh weight and dry weight of seedlings in the FG8-treated group were significantly higher than those in the control group. These results indicate that strain FG8 has both growth-promoting and biological control functions, which can provide a potential resource for the biological control of I. hawkeri leaf spot and the development of fungal fertilizers. Its field application effect and mechanism of action need to be further explored. Full article

Magnetic wall-climbing robots have great potential applications for the maintenance and inspection of large steel structures. However, they are susceptible to overturning when climbing over obstacles on vertical walls, primarily due to localized failures in the adhesion and shifts in the center of gravity. To address this issue, this paper presents an improved robot design featuring a passive adaptive tracked mobility module and a link-spring anti-overturning module. The adaptive tracked mobility module, incorporating spring tensioning mechanisms and belt press wheels, enables dynamic conformity to uneven walls and maintains stable magnetic adhesion. The link-spring anti-overturning module converts the front-end lift during obstacle crossing into an anti-overturning moment applied to the rear end of the robot. Notably, there is no need for additional drivers or control units. The structural design and three-dimensional modeling of the robot are carried out. Its working principle is analyzed, and parametric modeling and simulation analysis are performed. A physical prototype is developed and obstacle-crossing experiments are conducted on a vertical wall. The results demonstrate that the adaptive tracked mobility module and the anti-overturning module can successfully assist the robot in climbing over an obstacle with a maximum height of 23 mm, and the robot exhibits excellent stability while climbing over continuous obstacles and moving on uneven vertical walls. Full article

Complex psychosis (CP) remains one of the most challenging conditions in mental health, characterized by persistent symptoms, cognitive impairment, functional disability, and reduced autonomy. Traditional rehabilitation approaches, although essential, are often insufficient to address the multidimensional needs of these individuals. Over the past decade, rapid advances in digital health have opened new opportunities to enhance psychosocial rehabilitation, improve engagement, and personalize treatment pathways. This narrative review synthesizes current evidence on the use of digital and immersive technologies in the rehabilitation of people with CP, including virtual reality (VR), augmented reality (AR), telerehabilitation platforms, mobile health (m-Health) applications, digital phenotyping, and AI-assisted cognitive remediation. We examine clinical trials, feasibility studies, and real-world implementations published between 2015 and 2025, highlighting the efficacy of VR-based social cognition training, remote cognitive remediation, ecological momentary interventions, and hybrid digital–in-person rehabilitation models. Mechanisms of action, transfer to real-world functioning, and predictors of engagement are described. Barriers such as digital literacy, access disparities, privacy concerns, and clinical integration are critically discussed. We also outline future directions, including adaptive algorithms, biosensor integration, and the development of multimodal digital ecosystems tailored to individual recovery trajectories. By integrating technological innovation with recovery-oriented care, digital rehabilitation tools have the potential to transform the treatment landscape for people with CP. This review offers a roadmap for clinicians, researchers, and policymakers seeking to incorporate evidence-based digital solutions into modern psychiatric rehabilitation. Full article

This study proposes a heterogeneous fusion path planning framework for unmanned aerial vehicles (UAVs) operating in complex emergency rescue and civil environments. Existing single-mechanism metaheuristics—including Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), and Genetic Algorithms (GAs)—suffer from fundamental limitations in three-dimensional kinematic path planning: PSO converges rapidly but stagnates at local optima due to population variance collapse; ACO offers robust local exploitation but incurs prohibitive cold-start overhead; GAs maintain diversity at the cost of expensive crossover operations. To address these complementary deficiencies simultaneously, the proposed framework introduces a spherical coordinate representation that reduces computational complexity and naturally enforces UAV kinematic constraints, combined with adaptive weight factors and a serial PSO-ACO fusion strategy, and subsequently incorporates adaptive weight factors. A serial fusion strategy is then introduced, wherein the sub-optimal trajectory generated by the Spherical PSO phase is mapped into the ACO pheromone field via a Gaussian Kernel Density Mapping (GKDM) mechanism, enabling the ACO phase to perform fine-grained local exploitation within a kinematically feasible corridor. Various constraints along the flight path are formulated into distinct cost functions, which cover aircraft track length, pitch angle variation, altitude difference variation, obstacle avoidance, and smoothness; the core task of the algorithm is to find the flight path with the minimum total cost. The proposed algorithm is dedicated to UAV path planning in complex emergency rescue environments (disaster-stricken areas, hazardous zones) and is further applicable to civil low-altitude logistics delivery, industrial facility inspection, ecological environment monitoring and urban air mobility (UAM) scenarios with complex obstacle constraints. It can effectively improve the safety and efficiency of UAVs in reaching rescue points, delivering emergency supplies, conducting disaster surveys, and completing various civil low-altitude operation tasks. Full article

Multi-omnidirectional mobile robot formations offer significant advantages for applications in unstructured environments. However, under constraints such as limited field of view and high operator cognitive load, existing teleoperation frameworks struggle to guarantee formation safety and stability. In this study, a bilateral shared control framework for multi-robot formation that integrates intent perception and vortex-field modulation is proposed. First, an Intent-Mediated Asymmetric Vortex Modulation (IM-AVM) strategy is developed, where the operator’s micro-intentions are mapped to determine the topological orientation of a vortex field. By constructing a dynamic asymmetric modulation matrix, saddle points in the potential field are geometrically eliminated, enabling deadlock-free obstacle avoidance while maintaining a rigid formation. Second, a multi-dimensional perception-based dynamic authority arbitration and topological deadlock escape mechanism is constructed, facilitating a seamless transition from assisted deadlock to autonomous escape. Finally, a formation coordination system based on anisotropic flow field modulation and adaptive sliding mode control is designed. Rigid formation constraints are transformed into a tangential safe flow field, and robust tracking is subsequently achieved through an Adaptive Nonsingular Fast Terminal Sliding Mode Controller (ANFTSMC). Theoretical analysis and experimental results demonstrate that the proposed framework achieves collision-free navigation for the formation in simulated environments. Full article

Despite significant advancements in communication systems, inherent limitations persist in providing reliable data transmission for emerging applications with massive data exchanges. Semantic communication offers promising solutions by extracting and transmitting meaningful information rather than raw bit sequences. However, it faces challenges from high mobility and dynamic channel conditions in wireless environments. In this paper, we design a ground-to-air network architecture that integrates a rotary-wing unmanned aerial vehicle (UAV) and ground terminals to maximize semantic transmission efficiency while maintaining low energy consumption. This approach leverages the high mobility of the UAV for flexible deployment and the data reduction capabilities of semantic communication. Therefore, we formulate a multi-objective optimization problem to simultaneously balance the total semantic transmission rate and the UAV propulsion energy consumption by jointly optimizing the UAV hovering position, semantic encoding lengths, and resource block (RB) allocation. The problem is complex, with mixed continuous and discrete variables, which necessitates an advanced optimization method. To address these challenges, we propose a novel greedy-enhanced adaptive multi-objective cellular genetic algorithm (GEAMOCell), which utilizes an adaptive neighborhood selection mechanism to balance exploration and exploitation, and employs a crowding-guided archive feedback mechanism to maintain population diversity. The simulation results demonstrate that the proposed GEAMOCell algorithm outperforms baseline algorithms in terms of convergence, semantic transmission rate, and energy efficiency. Full article

Understanding and predicting how tourists move through a city is a challenging task, as it involves a complex interplay of spatial, temporal, and social factors. Traditional recommender systems often rely on structured data, trying to capture the nature of the problem. However, recent advances in Large Language Models (LLMs) open new possibilities for reasoning over richer, text-based representations of user context, even without a dedicated pre-training phase. In this study, we investigate the potential of LLMs to interpret and predict tourist movements in a real-world application scenario involving tourist visits to Verona, a municipality in Northern Italy, between 2014 and 2023. We propose an incremental prompt engineering approach that gradually enriches the model input, from spatial features alone to richer behavioral information, including visit histories, time information, and user cluster patterns. The approach is evaluated using six open-source models, enabling us to compare their accuracy and efficiency across various levels of contextual enrichment. The results provide a first insight about the abilities of LLMs to incorporate spatio-temporal contextual factors, thus improving predictions, while maintaining computational efficiency. The analysis of the model-generated explanations completes the picture by adding an interpretability dimension that most existing next-PoI prediction solutions lack. Overall, the study demonstrates the potential of LLMs to integrate multiple contextual dimensions in tourism mobility, highlighting the possibility of a more text-oriented, adaptive, and explainable T-RS. Full article