Internet Technology Letters最新文献

{"title":"A Multi-User Collaborative Recommendation Mechanism for Career Planning in Online Learning Over Edge Networks","authors":"Zhen Zhang,&nbsp;Guixin Luo,&nbsp;Jieyu Zhang","doi":"10.1002/itl2.70126","DOIUrl":"https://doi.org/10.1002/itl2.70126","url":null,"abstract":"<div>\u0000 \u0000 <p>In recent years, online learning platforms have gained popularity, particularly in the realm of career planning and skill development. However, most existing recommendation systems fail to fully integrate multi-behavioral user data and collaborative group preferences. This paper presents a Multi-User Collaborative Recommendation Mechanism for Career Planning in Online Learning (MCR-MCL), which combines multi-behavioral interaction data, group consensus modeling, and edge Networks to enhance personalized career planning recommendations. By leveraging edge network deployment, our system enables low-latency, localized updates that dynamically adapt to users' behaviors without frequent reliance on centralized cloud servers. We propose an innovative approach that leverages Graph Convolutional Networks (GCNs) to process user-item interactions and a behavioral independence modeling mechanism to avoid over-reliance on a single interaction type. We evaluate the effectiveness of the proposed mechanism using two real-world datasets—CareerMOOC and CareerEdNet—and demonstrate that our model significantly outperforms existing state-of-the-art methods in terms of recommendation accuracy, diversity, and low-latency adaptability through edge-based processing. The experimental results indicate that MCR-MCL can provide highly relevant, diverse, and dynamic recommendations that are essential for career planning in the context of online learning.</p>\u0000 </div>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5,"publicationDate":"2025-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998944","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Effective Centralized Power Control and Management of Nano-Grid Using DT Based Novel Distributed Framework","authors":"Jarabala Ranga,&nbsp;Gopinath Palai,&nbsp;Rabi N. Satpathy","doi":"10.1002/itl2.70080","DOIUrl":"https://doi.org/10.1002/itl2.70080","url":null,"abstract":"<div>\u0000 \u0000 <p>Nano-grid is an independent hybrid sustainable framework which uses both renewable and non-renewable power sources to continuously supply energy to load. Nano-grid finds its possibilities for the integration of distributed energy sources for realizing versatile and efficient energy management systems for future homes, local communities, and buildings. Nano-grid's energy trading system effectiveness might depend on various factors including core efficient management components such as energy storage systems (ESS) and renewable energy devices. Smart advanced functions in consumer devices and their unpredictable usage patterns result in unpredictable fluctuations in consumption of power. These fluctuations pose significant challenges in stability and quality of the power rid and create complex power imbalance issues which become harder to control and manage. Innovative power control and management models are essential to solve these issues in the nano-grid. In recent times, machine learning algorithms can be used to predict, track the current conditions, and make appropriate adjustments to the quality settings of power. In this research, effective centralized power control and management of the nano-grid using DT-based novel distributed framework is presented. This system utilizes a novel distributed framework based on a decision tree to enhance the agility and stability of complex and large-scale power systems. Performance measures like accuracy, F1-score, and RMSE are used to evaluate the performance of this system. This system will achieve better centralized power control and management.</p>\u0000 </div>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5,"publicationDate":"2025-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998946","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"EfficientDet-EdgeUAV: A Multi-Scale Fusion Architecture for Target Detection in UAV Imagery With Edge Computing Optimization","authors":"Chang Su,&nbsp;Xin Deng,&nbsp;Dehan Xue","doi":"10.1002/itl2.70118","DOIUrl":"https://doi.org/10.1002/itl2.70118","url":null,"abstract":"<div>\u0000 \u0000 <p>To overcome cloud computing's limitations—high latency and costly data transfers that hinder rapid UAV detection—plus the challenges of spotting tiny targets against complex backgrounds in wide-field aerial views, an edge computing solution with its specialized lightweight network: EfficientDet-EdgeUAV is proposed. The network employs a structurally optimized EfficientNet backbone through lightweight architecture modifications, integrating a Squeeze-and-Excitation attention mechanism to mitigate interference from complex backgrounds in target detection. The network architecture enhances small object detection by incorporating large-scale feature layers into the pyramid structure of the neck and applying lightweight architectural optimization to the neck module. The architecture further enhances detection robustness by implementing multi-scale feature fusion in the neck module, which strategically combines shallow-layer spatial details and deep-layer semantic representations to improve discernment of small objects with blurred boundaries. Through extensive experiments that comprehensively evaluate and validate the effectiveness of the proposed method, the experimental results demonstrate superior detection accuracy and efficiency on the VisDrone dataset compared to baseline and state-of-the-art methods. This demonstrates that the proposed method achieves exceptional effectiveness in real-time UAV imaging scenarios, providing critical technical references for civilian applications of drone technology in power line inspection, geological exploration, and search and rescue operations.</p>\u0000 </div>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5,"publicationDate":"2025-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144935027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing Car Tracking Systems With DNN-LoRa","authors":"Malak Abid Ali Khan,&nbsp;Senlin Luo","doi":"10.1002/itl2.70130","DOIUrl":"https://doi.org/10.1002/itl2.70130","url":null,"abstract":"<div>\u0000 \u0000 <p>This paper outlines a fusion of deep neural network and LoRa technology for car tracking optimization. LoRa's SX1301 gateway (GW) applies the Bayesian game parameter selection (BGPS) approach for switching the transmission power at the network server. At the same time, the car node (CN) uses a hybrid model to change the spreading factor and data rate. By reducing power losses among GWs, BGPS substantially increases the packet success rate (PSR) at the CN. The hybrid model enables adaptive decision-making, resulting in improved tracking precision and reduced latency with efficient energy usage. However, it exhibits a 95.9% PSR with increased latency noted at the lower bandwidth.</p>\u0000 </div>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5,"publicationDate":"2025-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144935026","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrating Big Data and AI for Network Security in 6G to Enhance University Financial Management","authors":"Jun Liang,&nbsp;Ling Pu,&nbsp; WeiweiSun","doi":"10.1002/itl2.70106","DOIUrl":"https://doi.org/10.1002/itl2.70106","url":null,"abstract":"<p>In the 6G network structures, the integration of Big Data (BD) and artificial intelligence (AI) is beneficial for the purpose of improving cybersecurity in university financial management systems. So, the integration of the BD and AI in the 6G structures are suggested in this study. Then, the conventional centralized security systems are ineffective in the rapid digitalization of financial transactions. Because these conventional systems are susceptible to single points of failure (SPF), delayed threat detection, and data privacy breaches. In the real-time (RT) financial backgrounds, these conventional systems face difficulties in protecting the network against advanced cyber threats. These situations will call for a decentralized, adaptive, and privacy-preserving (PP) security framework in the rapidly evolving 6G structures. This demanded framework may help in anomaly detection (AD) in financial transactions without affecting vital data. Thus, a novel federated learning (FL)-based AD in financial security (FL-AD-FS) framework is suggested in this study. To train the AI models collaboratively over several edge devices, this suggested model utilizes FL. This application will also ensure the privacy of the data. Then, in financial operations, the RT AD and threat mitigation was facilitated by the system, as it integrates with 6G-enabled (EC) edge computing. The simulations were conducted; from the outcomes, it is clear that the suggested FL-AD-FS model executes better by reducing false positive rates (FPRs), increasing detection (ACC) accuracy, and minimizing latency. In university backgrounds, secure, fast, and reliable monitoring of financial transactions was facilitated by this suggested method. For revolutionizing cybersecurity in digital financial systems, the potential of the integration of the FL, AI, and 6G technologies is demonstrated by the FL-AD-FS framework. For modern university financial management, this suggested method creates a customized scalable, secure, and privacy-aware solution.</p>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5,"publicationDate":"2025-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/itl2.70106","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144934929","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MAGAN-RT: A Lightweight Adversarial Style Transfer Network for Real-Time Cartoonization on Low-Power Edge Devices","authors":"Peng Guo","doi":"10.1002/itl2.70104","DOIUrl":"https://doi.org/10.1002/itl2.70104","url":null,"abstract":"<div>\u0000 \u0000 <p>Recent advances in neural style transfer (NST) and generative adversarial networks (GANs) have enabled photorealistic and artistic image stylization. However, deploying such models on resource-constrained edge devices remains challenging due to their high computational and memory demands. In this paper, we propose MAGAN-RT, a lightweight adversarial style transfer framework optimized for real-time cartoon-style transformation on low-power mobile and embedded platforms. MAGAN-RT integrates depthwise separable convolutions, inverted bottleneck residual blocks, and a multi-scale perceptual distillation strategy with auxiliary RGB supervision to enable efficient and expressive stylization. Furthermore, a real-image-based adversarial loss is employed to enhance realism while avoiding the artifacts commonly inherited from teacher models. Experimental results demonstrate that MAGAN-RT outperforms existing lightweight and mobile-compatible style transfer networks in both visual quality and runtime efficiency. It achieves state-of-the-art LPIPS, FID, and SSIM scores, while maintaining sub-10 ms inference latency on commercial smartphones, making it suitable for real-time applications such as mobile AR and video filters.</p>\u0000 </div>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5,"publicationDate":"2025-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144927598","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"UAV-Assisted C-RAN for On-Demand Cellular Coverage: Opportunities and Challenges","authors":"Byomakesh Mahapatra,&nbsp;Deepika Gupta,&nbsp;Pankaj Kumar Sharma","doi":"10.1002/itl2.70117","DOIUrl":"https://doi.org/10.1002/itl2.70117","url":null,"abstract":"<div>\u0000 \u0000 <p>The deployment of beyond fifth-generation (5G) infrastructure over disaster-affected regions, temporary hotspot situations (e.g., massive gatherings, etc.), and complex terrains (e.g., sea, hills, marshes, etc.) poses numerous challenges for cellular service providers. Recently, unmanned aerial vehicles (UAVs) have emerged as potential candidates to overcome the aforementioned technical issues based on their multi-role capabilities to serve as aerial base stations, mobile relays, and flying wireless access points. As such, the UAVs can act as portable platforms that can be deployed immediately on demand without requiring massive ground infrastructure to support wireless services. This article introduces the integration of UAVs to cloud radio access networks (C-RAN) for beyond 5G applications. Despite various advantages, limitations such as limited power backup, delicate hardware, and restricted payload make UAVs unsuitable for large-scale operations such as macro-base station. The article mainly focuses on the underlying opportunities and challenges to realize the UAV-assisted C-RAN (UC-RAN) architecture in view of three generic application scenarios, i.e., disaster management, hotspots, and complex terrains. A preliminary performance analysis via simulation is provided for the proposed UC-RAN architecture under the hotspot application scenario based on the relevant metrics.</p>\u0000 </div>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144927505","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Edge-Friendly NAS Framework for Mural Pattern Recognition via Structure-Aware Feature Fusion","authors":"Xianke Zhou,&nbsp;Wenjie Deng,&nbsp;Fengran Xie","doi":"10.1002/itl2.70109","DOIUrl":"https://doi.org/10.1002/itl2.70109","url":null,"abstract":"<div>\u0000 \u0000 <p>Ancient mural recognition faces unique challenges due to degradation, stylistic variations, and domain-specific symbolism. We propose a lightweight, edge-deployable neural architecture search (NAS) framework—SG-NAS-MPR—designed for accurate mural pattern recognition. Our framework integrates gated convolutions with frequency-domain fusion in a structure-aware module to enhance features under visual noise. A contrast-aware NAS strategy tailors compact backbones for real-time inference. Experiments on Dunhuang mural datasets show that our method surpasses existing CNN and NAS models in accuracy (93.4%) and F1-score (0.922), whereas reducing latency and model size. This work enables efficient and interpretable recognition in cultural heritage computing, supporting mobile museum applications and AR-based mural analysis.</p>\u0000 </div>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5,"publicationDate":"2025-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144923454","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IoT-Enabled Remote Teaching Management and Interaction: A Smart Framework for Real-Time Engagement and Environment Optimization","authors":"Shanshan Wang","doi":"10.1002/itl2.70123","DOIUrl":"https://doi.org/10.1002/itl2.70123","url":null,"abstract":"<div>\u0000 \u0000 <p>The shift to remote teaching has accelerated the demand for intelligent systems that can sustain high engagement and manage virtual classroom environments effectively. This paper proposes an IoT-enabled framework for remote teaching management and interaction that integrates environmental control, behavioral sensing, and real-time feedback mechanisms. The system adopts a three-tier architecture comprisinga perception layer with IoT sensors and edge computing nodes for data collection and preprocessing, a network layer that manages secure communication via the MQTT protocol, and an application layer offering cloud-based analytics, PID control, and user interfaces. This architecture enables precise regulation of temperature and lighting through PID control, as well as real-time tracking of student engagement using multimodal sensing and scoring algorithms. Extensive experiments involving 100 students and six comparison methods demonstrate the superiority of the proposed system in terms of engagement score, environmental stability (RMSE), delay, and instructor satisfaction. Quantitative metrics and visual analyses reveal that our solution reduces average data transmission latency to 21.4 ms and increases engagement by 12% over existing smart classroom models. These findings underscore the potential of IoT-driven intelligent frameworks in enhancing the interactivity, efficiency, and comfort of remote learning environments.</p>\u0000 </div>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5,"publicationDate":"2025-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144923455","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AI-Driven Secure 5G MIMO Enhancing Robotic Precision in Industrial and Service Applications","authors":"Yuanfang Wei","doi":"10.1002/itl2.70077","DOIUrl":"https://doi.org/10.1002/itl2.70077","url":null,"abstract":"<div>\u0000 \u0000 <p>The incorporation of AI in conjunction with secure 5G MIMO networks enhances the precision of consumer and industrial robotics. Ultra-reliable, low-latency communication paired with autonomous control enables faster, safer, and more accurate action execution in dynamic environments. However, contemporary robotic communication systems face challenges such as being highly susceptible to signal interference, network delays, cyber-attacks, and lack of adaptive capability. These obstacles particularly hinder remote control teleoperation and robotic efficiency in conditions which are highly volatile or constantly changing. The framework proposed, AI-Driven Secure 5G MIMO for Robotic Precision (AI-5G-MIMO-RP), uses AI adaptive signal processing to manage assistive cyber defense systems and strong 5G MIMO communications to overcome such challenges. MIMO technology not only increases data transmission speed, but also enhances dependability, while machine learning helps optimize data routing within the signals. AI-fortified cyber defenses detect and mitigate real-time and pre-emptive breaches, ensuring system communications cannot be tampered with. This approach supports application areas with smart precision like manufacturing, robotics for healthcare, facilitating automation in remote assistance, and serving in automated logistics. This technology enables dependably safe control and low-latency communication, guaranteeing accurate robot operation in complex tasks without human oversight. AI-5G-MIMO-RP creates a new standard in precision robotics control, network resilience, and operational efficiency. This technology reduces communication delays, increases network flexibility, and enhances system reliability, making industrial and service settings safer and more efficient than previous systems.</p>\u0000 </div>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5,"publicationDate":"2025-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144905454","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}