With the large-scale deployment of the Internet ofThings(IoT)devices,their weak securitymechanisms make them prime targets for malware attacks.Attackers often use Domain Generation Algorithm(DGA)to generate random dom...With the large-scale deployment of the Internet ofThings(IoT)devices,their weak securitymechanisms make them prime targets for malware attacks.Attackers often use Domain Generation Algorithm(DGA)to generate random domain names,hiding the real IP of Command and Control(C&C)servers to build botnets.Due to the randomness and dynamics of DGA,traditional methods struggle to detect them accurately,increasing the difficulty of network defense.This paper proposes a lightweight DGA detection model based on knowledge distillation for resource-constrained IoT environments.Specifically,a teacher model combining CharacterBERT,a bidirectional long short-term memory(BiLSTM)network,and attention mechanism(ATT)is constructed:it extracts character-level semantic features viaCharacterBERT,captures sequence dependencieswith the BiLSTM,and integrates theATT for key feature weighting,formingmulti-granularity feature fusion.An improved knowledge distillation approach transfers the teacher model’s learned knowledge to the simplified DistilBERT student model.Experimental results show the teacher model achieves 98.68%detection accuracy.The student modelmaintains slightly improved accuracy while significantly compressing parameters to approximately 38.4%of the teacher model’s scale,greatly reducing computational overhead for IoT deployment.展开更多
The exponential growth of Internet of Things(IoT)devices,autonomous systems,and digital services is generating massive volumes of big data,projected to exceed 291 zettabytes by 2027.Conventional cloud computing,despit...The exponential growth of Internet of Things(IoT)devices,autonomous systems,and digital services is generating massive volumes of big data,projected to exceed 291 zettabytes by 2027.Conventional cloud computing,despite its high processing and storage capacity,suffers from increased network latency,network congestion,and high operational costs,making it unsuitable for latency-sensitive applications.Edge computing addresses these issues by processing data near the source but faces scalability challenges and elevated Total Cost of Ownership(TCO).Hybrid solutions,such as fog computing,cloudlets,and Mobile Edge Computing(MEC),attempt to balance cost and performance;however,they still struggle with limited resource sharing and high deployment expenses.This paper proposes Public Edge as a Service(PEaaS),a novel paradigm that utilizes idle resources contributed by universities,enterprises,cellular operators,and individuals under a collaborative service model.By decentralizing computation and enabling multi-tenant resource sharing,PEaaS reduces reliance on centralized cloud infrastructure,minimizes communication costs,and enhances scalability.The proposed framework is evaluated using EdgeCloudSim under varying workloads,for keymetrics such as latency,communication cost,server utilization,and task failure rate.Results reveal that while cloud has a task failure rate rising sharply to 12.3%at 2000 devices,PEaaS maintains a low rate of 2.5%,closely matching edge computing.Furthermore,communication costs remain 25% lower than cloud and latency remains below 0.3,even under peak load.These findings demonstrate that PEaaS achieves near-edge performance with reduced costs and enhanced scalability,offering a sustainable and economically viable solution for next-generation computing environments.展开更多
With the rapid development of power Internet of Things(IoT)scenarios such as smart factories and smart homes,numerous intelligent terminal devices and real-time interactive applications impose higher demands on comput...With the rapid development of power Internet of Things(IoT)scenarios such as smart factories and smart homes,numerous intelligent terminal devices and real-time interactive applications impose higher demands on computing latency and resource supply efficiency.Multi-access edge computing technology deploys cloud computing capabilities at the network edge;constructs distributed computing nodes and multi-access systems and offers infrastructure support for services with low latency and high reliability.Existing research relies on a strong assumption that the environmental state is fully observable and fails to thoroughly consider the continuous time-varying features of edge server load fluctuations,leading to insufficient adaptability of the model in a heterogeneous dynamic environment.Thus,this paper establishes a framework for end-edge collaborative task offloading based on a partially observable Markov decision-making process(POMDP)and proposes a method for end-edge collaborative task offloading in heterogeneous scenarios.It achieves time-series modeling of the historical load characteristics of edge servers and endows the agent with the ability to be aware of the load in dynamic environmental states.Moreover,by dynamically assessing the exploration value of historical trajectories in the central trajectory pool and adjusting the sample weight distribution,directional exploration and strategy optimization of high-value trajectories are realized.Experimental results indicate that the proposed method exhibits distinct advantages compared with existing methods in terms of average delay and task failure rate and also verifies the method’s robustness in a dynamic environment.展开更多
Coulomb drag refers to the phenomenon in which a current driven through one conducting layer induces a voltage nearby,electrically isolated layer sorely through interlayer Coulomb interactions between charge carriers....Coulomb drag refers to the phenomenon in which a current driven through one conducting layer induces a voltage nearby,electrically isolated layer sorely through interlayer Coulomb interactions between charge carriers.It has been extensively studied in various systems,including parallel nanowires,double quantum wells,and double-layer graphene.Here,we report the observation of Coulomb drag in a novel system consisting of two graphene layers separated laterally by a 30 nm gap within the material plane,exhibiting behavior distinct from that in vertical graphene heterostructures.Our experiments reveal pronounced negative drag resistances under an out-of-plane magnetic field at the quantum Hall edges,reaching a maximum when the carrier densities in both graphene layers are tuned to the charge neutrality point via gate voltages.Our work establish two separate and spatially closed quantum Hall edge modes as a new platform to explore electronic interaction physics between one dimensional systems.展开更多
Given the growing number of vehicle accidents caused by unintended acceleration and braking failure,verifying Sudden Unintended Acceleration(SUA)incidents has become a persistent challenge.A central issue of debate is...Given the growing number of vehicle accidents caused by unintended acceleration and braking failure,verifying Sudden Unintended Acceleration(SUA)incidents has become a persistent challenge.A central issue of debate is whether such events stem frommechanical malfunctions or driver pedalmisapplications.However,existing verification procedures implemented by vehiclemanufacturers often involve closed tests after vehicle recalls;thus raising ongoing concerns about reliability and transparency.Consequently,there is a growing need for a user-driven framework that enables independent data acquisition and verification.Although previous studies have addressed SUA detection using deep learning,few have explored howclass granularity optimization affects power efficiency and inference performance in real-time Edge AI systems.To address this problem,this work presents a cloud-assisted artificial intelligence(AI)solution for the reliable verification of SUA occurrences.The proposed system integrates multimodal sensor streams including camera-based foot images,On-Board Diagnostics II(OBD-II)signals,and six-axismeasurements to determine whether the brake pedal was actually engaged at themoment of a suspected SUA.Beyond image acquisition,convolutional neural network(CNN)models perform real-time inference to classify the driver’s pedal operation states with the resulting outputs transmitted and archived in the cloud.A dedicated dataset of brake and accelerator pedal images was collected from 15 vehicles produced by 6 domestic and international manufacturers.Using this dataset,transfer learning techniques were applied to compare and analyze model performance and generalization as the CNN class granularity varied from coarse to fine levels.Furthermore,classification performance was evaluated in terms of latency and power efficiency under different class configurations.The experimental results demonstrated that the proposed solution identified the driver’s pedal behavior accurately and promptly,with the two-class model achieving the highest F1-score and accuracy among all granularity settings.展开更多
High-elevation forests are among the most climate-sensitive ecosystems,and understanding their growth responses is crucial for predicting ecological consequences under future climate change.The climate sensitivity of ...High-elevation forests are among the most climate-sensitive ecosystems,and understanding their growth responses is crucial for predicting ecological consequences under future climate change.The climate sensitivity of tree species in the Hyrcanian forests in the Alborz Mountains of northern Iran,one of the southernmost temperate deciduous forests in the Northern Hemisphere,remains largely unexplored.In particular,Acer hyrcanum Fisch.&C.A.Mey.,growing mainly at high elevations,has not yet been studied in detail in dendroclimatology.Here,we present the first tree-ring chronology of Acer hyrcanum spanning 1814-2022 and analyze its growth-climate relationships to assess how this species reflects climatic sensitivity at the upper forest limit.The results reveal significant positive correlations between tree-ring width and temperature,particularly from May to September,suggesting that warmer growing-season temperatures enhance tree growth.In contrast,tree-ring width showed negative correlations with precipitation and standardized precipitation-evapotranspiration index,especially from January to May,and with cloud cover from March to May.These findings suggest that moisture availability does not limit radial growth in Acer hyrcanum and that the precipitation and water surplus signals may instead reflect the influence of cloud cover,which reduces sunlight availability during critical early-season months.This study contributes to the growing body of dendroclimatic research in the Alborz Mountains and,more broadly,on Acer species,particularly in high-elevation ecosystems where such studies are scarce.It also provides valuable insights into how Acer hyrcanum may respond to future climate change.展开更多
This work presents a systematic analysis of proton-induced total ionizing dose(TID)effects in 1.2 k V silicon carbide(SiC)power devices with various edge termination structures.Three edge terminations including ring-a...This work presents a systematic analysis of proton-induced total ionizing dose(TID)effects in 1.2 k V silicon carbide(SiC)power devices with various edge termination structures.Three edge terminations including ring-assisted junction termination extension(RA-JTE),multiple floating zone JTE(MFZ-JTE),and field limiting rings(FLR)were fabricated and irradiated with45 Me V protons at fluences ranging from 1×10^(12) to 1×10^(14) cm^(-2).Experimental results,supported by TCAD simulations,show that the RA-JTE structure maintained stable breakdown performance with less than 1%variation due to its effective electric field redistribution by multiple P+rings.In contrast,MFZ-JTE and FLR exhibit breakdown voltage shifts of 6.1%and 15.2%,respectively,under the highest fluence.These results demonstrate the superior radiation tolerance of the RA-JTE structure under TID conditions and provide practical design guidance for radiation-hardened Si C power devices in space and other highradiation environments.展开更多
Multimodal spatiotemporal data from smart city consumer electronics present critical challenges including cross-modal temporal misalignment,unreliable data quality,limited joint modeling of spatial and temporal depend...Multimodal spatiotemporal data from smart city consumer electronics present critical challenges including cross-modal temporal misalignment,unreliable data quality,limited joint modeling of spatial and temporal dependencies,and weak resilience to adversarial updates.To address these limitations,EdgeST-Fusion is introduced as a cross-modal federated graph transformer framework for context-aware smart city analytics.The architecture integrates cross-modal embedding networks for modality alignment,graph transformer encoders for spatial dependency modeling,temporal self-attention for dynamic pattern learning,and adaptive anomaly detection to ensure data quality and security during aggregation.A privacy-preserving federated learning protocol with differential privacy guarantees enables collaborative model training without centralizing sensitive data.The framework employs data-quality-aware weighted aggregation to enhance robustness against noisy and malicious client updates.Experimental evaluation on the GeoLife,PeMS-Bay,and SmartHome+datasets demonstrates that EdgeST-Fusion achieves 21.8%improvement in prediction accuracy,35.7%reduction in communication overhead,and 29.4%enhancement in security resilience compared to recent baselines.Real-world deployment across three smart city testbeds validates practical viability with 90.0%average accuracy and sub-250 ms inference latency.The proposed framework remains feasible for deployment on heterogeneous and resource-constrained consumer electronics devices whilemaintaining strong privacy guarantees and scalability for large-scale urban environments.展开更多
Traffic sign detection is a critical component of driving systems.Single-stage network-based traffic sign detection algorithms,renowned for their fast detection speeds and high accuracy,have become the dominant approa...Traffic sign detection is a critical component of driving systems.Single-stage network-based traffic sign detection algorithms,renowned for their fast detection speeds and high accuracy,have become the dominant approach in current practices.However,in complex and dynamic traffic scenes,particularly with smaller traffic sign objects,challenges such as missed and false detections can lead to reduced overall detection accuracy.To address this issue,this paper proposes a detection algorithm that integrates edge and shape information.Recognizing that traffic signs have specific shapes and distinct edge contours,this paper introduces an edge feature extraction branch within the backbone network,enabling adaptive fusion with features of the same hierarchical level.Additionally,a shape prior convolution module is designed to replaces the first two convolutional modules of the backbone network,aimed at enhancing the model's perception ability for specific shape objects and reducing its sensitivity to background noise.The algorithm was evaluated on the CCTSDB and TT100k datasets,and compared to YOLOv8s,the mAP50 values increased by 3.0%and 10.4%,respectively,demonstrating the effectiveness of the proposed method in improving the accuracy of traffic sign detection.展开更多
Due to the growth of smart cities,many real-time systems have been developed to support smart cities using Internet of Things(IoT)and emerging technologies.They are formulated to collect the data for environment monit...Due to the growth of smart cities,many real-time systems have been developed to support smart cities using Internet of Things(IoT)and emerging technologies.They are formulated to collect the data for environment monitoring and automate the communication process.In recent decades,researchers have made many efforts to propose autonomous systems for manipulating network data and providing on-time responses in critical operations.However,the widespread use of IoT devices in resource-constrained applications and mobile sensor networks introduces significant research challenges for cybersecurity.These systems are vulnerable to a variety of cyberattacks,including unauthorized access,denial-of-service attacks,and data leakage,which compromise the network’s security.Additionally,uneven load balancing between mobile IoT devices,which frequently experience link interferences,compromises the trustworthiness of the system.This paper introduces a Multi-Agent secured framework using lightweight edge computing to enhance cybersecurity for sensor networks,aiming to leverage artificial intelligence for adaptive routing and multi-metric trust evaluation to achieve data privacy and mitigate potential threats.Moreover,it enhances the efficiency of distributed sensors for energy consumption through intelligent data analytics techniques,resulting in highly consistent and low-latency network communication.Using simulations,the proposed framework reveals its significant performance compared to state-of-the-art approaches for energy consumption by 43%,latency by 46%,network throughput by 51%,packet loss rate by 40%,and denial of service attacks by 42%.展开更多
The Hall effect of elastic waves has attracted much attention due to its unique properties.A hexagonal lattice phononic crystal plate model is designed in this paper.By changing the spatial symmetry of the unit cell,a...The Hall effect of elastic waves has attracted much attention due to its unique properties.A hexagonal lattice phononic crystal plate model is designed in this paper.By changing the spatial symmetry of the unit cell,a band gap for the A_(0) Lamb wave is opened.The existence of the edge state of the phononic crystal plate is obtained by finite element simulation.It is found that both zigzag-type edge and bridge edge are topological edge states by analysis of the band structure of the supercell.A rectangular model with a straight channel is designed and the simulation results show that the two types of channels are topologically protected only for the A_(0) mode Lamb wave but not for the S_(0) mode.In addition,the results of numerical simulation are verified by experimental data measured by a laser vibrometer.Finally,it is found that neither upside V-shaped channels nor channels with defects will affect the stable propagation of A_(0) Lamb waves along the proposed route.This proposed model and method are helpful in broadening the means of regulating elastic waves in phononic crystal structures,and extending practical application of topological edge states in such structures.展开更多
With the rapid expansion of drone applications,accurate detection of objects in aerial imagery has become crucial for intelligent transportation,urban management,and emergency rescue missions.However,existing methods ...With the rapid expansion of drone applications,accurate detection of objects in aerial imagery has become crucial for intelligent transportation,urban management,and emergency rescue missions.However,existing methods face numerous challenges in practical deployment,including scale variation handling,feature degradation,and complex backgrounds.To address these issues,we propose Edge-enhanced and Detail-Capturing You Only Look Once(EHDC-YOLO),a novel framework for object detection in Unmanned Aerial Vehicle(UAV)imagery.Based on the You Only Look Once version 11 nano(YOLOv11n)baseline,EHDC-YOLO systematically introduces several architectural enhancements:(1)a Multi-Scale Edge Enhancement(MSEE)module that leverages multi-scale pooling and edge information to enhance boundary feature extraction;(2)an Enhanced Feature Pyramid Network(EFPN)that integrates P2-level features with Cross Stage Partial(CSP)structures and OmniKernel convolutions for better fine-grained representation;and(3)Dynamic Head(DyHead)with multi-dimensional attention mechanisms for enhanced cross-scale modeling and perspective adaptability.Comprehensive experiments on the Vision meets Drones for Detection(VisDrone-DET)2019 dataset demonstrate that EHDC-YOLO achieves significant improvements,increasing mean Average Precision(mAP)@0.5 from 33.2%to 46.1%(an absolute improvement of 12.9 percentage points)and mAP@0.5:0.95 from 19.5%to 28.0%(an absolute improvement of 8.5 percentage points)compared with the YOLOv11n baseline,while maintaining a reasonable parameter count(2.81 M vs the baseline’s 2.58 M).Further ablation studies confirm the effectiveness of each proposed component,while visualization results highlight EHDC-YOLO’s superior performance in detecting objects and handling occlusions in complex drone scenarios.展开更多
In real-world autonomous driving tests,unexpected events such as pedestrians or wild animals suddenly entering the driving path can occur.Conducting actual test drives under various weather conditions may also lead to...In real-world autonomous driving tests,unexpected events such as pedestrians or wild animals suddenly entering the driving path can occur.Conducting actual test drives under various weather conditions may also lead to dangerous situations.Furthermore,autonomous vehicles may operate abnormally in bad weather due to limitations of their sensors and GPS.Driving simulators,which replicate driving conditions nearly identical to those in the real world,can drastically reduce the time and cost required for market entry validation;consequently,they have become widely used.In this paper,we design a virtual driving test environment capable of collecting and verifying SiLS data under adverse weather conditions using multi-source images.The proposed method generates a virtual testing environment that incorporates various events,including weather,time of day,and moving objects,that cannot be easily verified in real-world autonomous driving tests.By setting up scenario-based virtual environment events,multi-source image analysis and verification using real-world DCUs(Data Concentrator Units)with V2X-Car edge cloud can effectively address risk factors that may arise in real-world situations.We tested and validated the proposed method with scenarios employing V2X communication and multi-source image analysis.展开更多
There has been an increasing emphasis on performing deep neural network(DNN)inference locally on edge devices due to challenges such as network congestion and security concerns.However,as DRAM process technology conti...There has been an increasing emphasis on performing deep neural network(DNN)inference locally on edge devices due to challenges such as network congestion and security concerns.However,as DRAM process technology continues to scale down,the bit-flip errors in the memory of edge devices become more frequent,thereby leading to substantial DNN inference accuracy loss.Though several techniques have been proposed to alleviate the accuracy loss in edge environments,they require complex computations and additional parity bits for error correction,thus resulting in significant performance and storage overheads.In this paper,we propose FeatherGuard,a data-driven lightweight error protection scheme for DNN inference on edge devices.FeatherGuard selectively protects critical bit positions(that have a significant impact on DNN inference accuracy)against bit-flip errors,by considering various DNN characteristics(e.g.,data format,layer-wise weight distribution,actually stored logical values).Thus,it achieves high error tolerability during DNN inference.Since FeatherGuard reduces the bit-flip errors based on only a few simple arithmetic operations(e.g.,NOT operations)without parity bits,it causes negligible performance overhead and no storage overhead.Our experimental results show that FeatherGuard improves the error tolerability by up to 6667×and 4000×,compared to the conventional systems and the state-of-the-art error protection technique for edge environments,respectively.展开更多
Industrial operators need reliable communication in high-noise,safety-critical environments where speech or touch input is often impractical.Existing gesture systems either miss real-time deadlines on resourceconstrai...Industrial operators need reliable communication in high-noise,safety-critical environments where speech or touch input is often impractical.Existing gesture systems either miss real-time deadlines on resourceconstrained hardware or lose accuracy under occlusion,vibration,and lighting changes.We introduce Industrial EdgeSign,a dual-path framework that combines hardware-aware neural architecture search(NAS)with large multimodalmodel(LMM)guided semantics to deliver robust,low-latency gesture recognition on edge devices.The searched model uses a truncated ResNet50 front end,a dimensional-reduction network that preserves spatiotemporal structure for tubelet-based attention,and localized Transformer layers tuned for on-device inference.To reduce reliance on gloss annotations and mitigate domain shift,we distill semantics from factory-tuned vision-language models and pre-train with masked language modeling and video-text contrastive objectives,aligning visual features with a shared text space.OnML2HP and SHREC’17,theNAS-derived architecture attains 94.7% accuracywith 86ms inference latency and about 5.9W power on Jetson Nano.Under occlusion,lighting shifts,andmotion blur,accuracy remains above 82%.For safetycritical commands,the emergency-stop gesture achieves 72 ms 99th percentile latency with 99.7% fail-safe triggering.Ablation studies confirm the contribution of the spatiotemporal tubelet extractor and text-side pre-training,and we observe gains in translation quality(BLEU-422.33).These results show that Industrial EdgeSign provides accurate,resource-aware,and safety-aligned gesture recognition suitable for deployment in smart factory settings.展开更多
As a fundamental component in computer vision,edges can be categorized into four types based on discontinuities in reflectance,illumination,surface normal,or depth.While deep CNNs have significantly advanced generic e...As a fundamental component in computer vision,edges can be categorized into four types based on discontinuities in reflectance,illumination,surface normal,or depth.While deep CNNs have significantly advanced generic edge detection,real-time multi-class semantic edge detection under resource constraints remains challenging.To address this,we propose a lightweight framework based on PiDiNet that enables fine-grained semantic edge detection.Our model simultaneously predicts background and four edge categories from full-resolution inputs,balancing accuracy and efficiency.Key contributions include:a multi-channel output structure expanding binary edge prediction to five classes,supported by a deep supervision mechanism;a dynamic class-balancing strategy combining adaptive weighting with physical priors to handle extreme class imbalance;and maintained architectural efficiency enabling real-time inference.Extensive evaluations on BSDS-RIND show our approach achieves accuracy competitive with state-of-the-art methods while operating in real time.展开更多
The intelligent environmental sensing systems are quickly transforming the sparse and retrospective monitoring to dense and decision-oriented environmental intelligence.This review brings together the manner in which ...The intelligent environmental sensing systems are quickly transforming the sparse and retrospective monitoring to dense and decision-oriented environmental intelligence.This review brings together the manner in which integration of Internet of Things(IoT)sensing,edge computing,and real-time analytics facilitates timely detection,interpretation,and prediction of the environmental conditions across the applications,such as urban air quality,watershed and coastal surveillance,industrial safety,agriculture,and disaster response.We define end-to-end architectural patterns to organize devices,edge nodes,and cloud services to satisfy latency,reliability,bandwidth,and governance constraints with emphasis on event-time processing,adaptive offloading,and hierarchical aggregation.Then we look at sensing and infrastructure foundations,emphasizing the effects of sensor modality and power autonomy,connectivity,and the practices of calibration on the practicable analytics and eventual plausibility.It is on this basis that we examine real-time analytics pipelines and Artificial Intelligence(AI)techniques to preprocess,sensor combine,anomaly detect,and short-horizon forecast,with a focus on edge-deployable models,quantification of uncertainties,and query resistance to drift and domain shift.Lastly,we address the realities of deployment that condition operational success,such as lifecycle engineering,provenance-aware data management,security and privacy risks,ethical governance,and evaluation methodologies,which place end-to-end latency and field generalization as a priority.This review offers cohesion to algorithmic capabilities and systems engineering and governance to define an overall framework,show open areas of research directions,and provide practical recommendations on how to design trustworthy,scalable,and sustainable environmental monitoring systems.展开更多
The personalized fine-tuning of large languagemodels(LLMs)on edge devices is severely constrained by limited computation resources.Although split federated learning alleviates on-device burdens,its effectiveness dimin...The personalized fine-tuning of large languagemodels(LLMs)on edge devices is severely constrained by limited computation resources.Although split federated learning alleviates on-device burdens,its effectiveness diminishes in few-shot reasoning scenarios due to the low data efficiency of conventional supervised fine-tuning,which leads to excessive communication overhead.To address this,we propose Language-Empowered Split Fine-Tuning(LESFT),a framework that integrates split architectures with a contrastive-inspired fine-tuning paradigm.LESFT simultaneously learns frommultiple logically equivalent but linguistically diverse reasoning chains,providing richer supervisory signals and improving data efficiency.This process-oriented training allows more effective reasoning adaptation with fewer samples.Extensive experiments demonstrate that LESFT consistently outperforms strong baselines such as SplitLoRA in task accuracy.LESFT consistently outperforms strong baselines on GSM8K,CommonsenseQA,and AQUA_RAT,with the largest gains observed on Qwen2.5-3B.These results indicate that LESFT can effectively adapt large language models for reasoning tasks under the computational and communication constraints of edge environments.展开更多
Binary segmentation tasks in computer vision exhibit diverse appearance distributions and complex boundary characteristics.To address the limited generalization and adaptability of existing models across heterogeneous...Binary segmentation tasks in computer vision exhibit diverse appearance distributions and complex boundary characteristics.To address the limited generalization and adaptability of existing models across heterogeneous tasks,we propose Abel-Net,an Aggregated Bilateral Edge Localization Network designed as a universal framework formulti-task binary segmentation.Abel-Net integrates global and local contextual cues to enhance feature learning and edge precision.Specifically,amulti-scale feature pyramid fusion strategy is implemented via anAggregated Skip Connection(ASC)module to strengthen feature adaptability,while the Edge Dual Localization(EDL)mechanism performs coarse-to-fine refinement through edge-aware supervision.Additionally,Edge Attention(EA)and Edge Fusion Attention(EFA)modules prioritize edge-critical regions and facilitate accurate boundary alignment.Extensive experiments on nine diverse binary segmentation tasks demonstrate thatAbel-Net performs comparably to or surpasses state-of-the-art task-specific networks,exhibiting strong adaptability to a wide range of visual perception challenges.展开更多
The intelligent transportation systems require secure,low-latency,and reliable communication architectures to enable the real-time vehicular application.This paper proposes an edge-intelligent semantic aggregation(EIS...The intelligent transportation systems require secure,low-latency,and reliable communication architectures to enable the real-time vehicular application.This paper proposes an edge-intelligent semantic aggregation(EISA)framework for 6G unmanned aerial vehicle(UAV)-assisted Internet of vehicles(IoV)networks that integrates task-driven semantic communication,deep reinforcement learning(DRL)-based edge intelligence,and blockchain-based semantic validation across 6G terahertz(THz)links.UAVs in the proposed architecture serve as adaptive edge nodes that receive semantically vital information about the vehicle at any given stage,optimize aggregation and transmission parameters dynamically,and guarantee data integrity through a structured,lightweight consortium blockchain that signs semantically detailed representations rather than raw packets.Simulation results from a hybrid NS-3,MATLAB,and Python environment indicate that the proposed framework can achieve up to 45%reduction in end-to-end latency,an approximately 70%increase in throughput,and semantic efficiency with blockchain verification delays of less than 20 ms(more than 98%).These findings support the effectiveness of the proposed co-design for achieving context-aware,energy-efficient,and reliable communication under heavy-traffic conditions.The proposed framework provides a flexible and scalable foundation for next-generation 6G-enabled automotive networks,with subsequent growth toward federated learning-based collaborative intelligence,digital-twinassisted traffic modeling,and quantum-safe blockchain mechanisms to enhance scalability,intelligence,and long-term security.展开更多
基金supported by the following projects:National Natural Science Foundation of China(62461041)Natural Science Foundation of Jiangxi Province China(20242BAB25068).
文摘With the large-scale deployment of the Internet ofThings(IoT)devices,their weak securitymechanisms make them prime targets for malware attacks.Attackers often use Domain Generation Algorithm(DGA)to generate random domain names,hiding the real IP of Command and Control(C&C)servers to build botnets.Due to the randomness and dynamics of DGA,traditional methods struggle to detect them accurately,increasing the difficulty of network defense.This paper proposes a lightweight DGA detection model based on knowledge distillation for resource-constrained IoT environments.Specifically,a teacher model combining CharacterBERT,a bidirectional long short-term memory(BiLSTM)network,and attention mechanism(ATT)is constructed:it extracts character-level semantic features viaCharacterBERT,captures sequence dependencieswith the BiLSTM,and integrates theATT for key feature weighting,formingmulti-granularity feature fusion.An improved knowledge distillation approach transfers the teacher model’s learned knowledge to the simplified DistilBERT student model.Experimental results show the teacher model achieves 98.68%detection accuracy.The student modelmaintains slightly improved accuracy while significantly compressing parameters to approximately 38.4%of the teacher model’s scale,greatly reducing computational overhead for IoT deployment.
文摘The exponential growth of Internet of Things(IoT)devices,autonomous systems,and digital services is generating massive volumes of big data,projected to exceed 291 zettabytes by 2027.Conventional cloud computing,despite its high processing and storage capacity,suffers from increased network latency,network congestion,and high operational costs,making it unsuitable for latency-sensitive applications.Edge computing addresses these issues by processing data near the source but faces scalability challenges and elevated Total Cost of Ownership(TCO).Hybrid solutions,such as fog computing,cloudlets,and Mobile Edge Computing(MEC),attempt to balance cost and performance;however,they still struggle with limited resource sharing and high deployment expenses.This paper proposes Public Edge as a Service(PEaaS),a novel paradigm that utilizes idle resources contributed by universities,enterprises,cellular operators,and individuals under a collaborative service model.By decentralizing computation and enabling multi-tenant resource sharing,PEaaS reduces reliance on centralized cloud infrastructure,minimizes communication costs,and enhances scalability.The proposed framework is evaluated using EdgeCloudSim under varying workloads,for keymetrics such as latency,communication cost,server utilization,and task failure rate.Results reveal that while cloud has a task failure rate rising sharply to 12.3%at 2000 devices,PEaaS maintains a low rate of 2.5%,closely matching edge computing.Furthermore,communication costs remain 25% lower than cloud and latency remains below 0.3,even under peak load.These findings demonstrate that PEaaS achieves near-edge performance with reduced costs and enhanced scalability,offering a sustainable and economically viable solution for next-generation computing environments.
基金funded by the State Grid Corporation Science and Technology Project“Research and Application of Key Technologies for Integrated Sensing and Computing for Intelligent Operation of Power Grid”(Grant No.5700-202318596A-3-2-ZN).
文摘With the rapid development of power Internet of Things(IoT)scenarios such as smart factories and smart homes,numerous intelligent terminal devices and real-time interactive applications impose higher demands on computing latency and resource supply efficiency.Multi-access edge computing technology deploys cloud computing capabilities at the network edge;constructs distributed computing nodes and multi-access systems and offers infrastructure support for services with low latency and high reliability.Existing research relies on a strong assumption that the environmental state is fully observable and fails to thoroughly consider the continuous time-varying features of edge server load fluctuations,leading to insufficient adaptability of the model in a heterogeneous dynamic environment.Thus,this paper establishes a framework for end-edge collaborative task offloading based on a partially observable Markov decision-making process(POMDP)and proposes a method for end-edge collaborative task offloading in heterogeneous scenarios.It achieves time-series modeling of the historical load characteristics of edge servers and endows the agent with the ability to be aware of the load in dynamic environmental states.Moreover,by dynamically assessing the exploration value of historical trajectories in the central trajectory pool and adjusting the sample weight distribution,directional exploration and strategy optimization of high-value trajectories are realized.Experimental results indicate that the proposed method exhibits distinct advantages compared with existing methods in terms of average delay and task failure rate and also verifies the method’s robustness in a dynamic environment.
基金support from the National Key Projects for Research and Development of China(Grant Nos.2022YFA1204700,2021YFA1400400)National Natural Science Foundation of China(Grant No.12525403)+3 种基金Natural Science Foundation of Jiangsu Province(Grant Nos.BK20220066,BK20233001)Program for Innovative Talents and Entrepreneur in Jiangsu(Grant No.JSSCTD202101)support from the JSPS KAKENHI(Grant Numbers 21H05233 and 23H02052)World Premier International Research Center Initiative(WPI),MEXT,Japan.
文摘Coulomb drag refers to the phenomenon in which a current driven through one conducting layer induces a voltage nearby,electrically isolated layer sorely through interlayer Coulomb interactions between charge carriers.It has been extensively studied in various systems,including parallel nanowires,double quantum wells,and double-layer graphene.Here,we report the observation of Coulomb drag in a novel system consisting of two graphene layers separated laterally by a 30 nm gap within the material plane,exhibiting behavior distinct from that in vertical graphene heterostructures.Our experiments reveal pronounced negative drag resistances under an out-of-plane magnetic field at the quantum Hall edges,reaching a maximum when the carrier densities in both graphene layers are tuned to the charge neutrality point via gate voltages.Our work establish two separate and spatially closed quantum Hall edge modes as a new platform to explore electronic interaction physics between one dimensional systems.
基金supported by Basic Science Research Program to Research Institute for Basic Sciences(RIBS)of Jeju National University through the National Research Foundation of Korea(NRF)funded by the Ministry of Education(RS-2019-NR040080)This research was also carried out with the support of the Jeju RISE Center,funded by the Ministry of Education and Jeju Special Self-Governing Province in 2025,as part of the“Regional Innovation System&Education(RISE):Glocal University 30”initiative.
文摘Given the growing number of vehicle accidents caused by unintended acceleration and braking failure,verifying Sudden Unintended Acceleration(SUA)incidents has become a persistent challenge.A central issue of debate is whether such events stem frommechanical malfunctions or driver pedalmisapplications.However,existing verification procedures implemented by vehiclemanufacturers often involve closed tests after vehicle recalls;thus raising ongoing concerns about reliability and transparency.Consequently,there is a growing need for a user-driven framework that enables independent data acquisition and verification.Although previous studies have addressed SUA detection using deep learning,few have explored howclass granularity optimization affects power efficiency and inference performance in real-time Edge AI systems.To address this problem,this work presents a cloud-assisted artificial intelligence(AI)solution for the reliable verification of SUA occurrences.The proposed system integrates multimodal sensor streams including camera-based foot images,On-Board Diagnostics II(OBD-II)signals,and six-axismeasurements to determine whether the brake pedal was actually engaged at themoment of a suspected SUA.Beyond image acquisition,convolutional neural network(CNN)models perform real-time inference to classify the driver’s pedal operation states with the resulting outputs transmitted and archived in the cloud.A dedicated dataset of brake and accelerator pedal images was collected from 15 vehicles produced by 6 domestic and international manufacturers.Using this dataset,transfer learning techniques were applied to compare and analyze model performance and generalization as the CNN class granularity varied from coarse to fine levels.Furthermore,classification performance was evaluated in terms of latency and power efficiency under different class configurations.The experimental results demonstrated that the proposed solution identified the driver’s pedal behavior accurately and promptly,with the two-class model achieving the highest F1-score and accuracy among all granularity settings.
基金supported by the Alexander von Humboldt Foundation(AvH),which provided a research stay for HM(Humboldt-ID number 1222705).
文摘High-elevation forests are among the most climate-sensitive ecosystems,and understanding their growth responses is crucial for predicting ecological consequences under future climate change.The climate sensitivity of tree species in the Hyrcanian forests in the Alborz Mountains of northern Iran,one of the southernmost temperate deciduous forests in the Northern Hemisphere,remains largely unexplored.In particular,Acer hyrcanum Fisch.&C.A.Mey.,growing mainly at high elevations,has not yet been studied in detail in dendroclimatology.Here,we present the first tree-ring chronology of Acer hyrcanum spanning 1814-2022 and analyze its growth-climate relationships to assess how this species reflects climatic sensitivity at the upper forest limit.The results reveal significant positive correlations between tree-ring width and temperature,particularly from May to September,suggesting that warmer growing-season temperatures enhance tree growth.In contrast,tree-ring width showed negative correlations with precipitation and standardized precipitation-evapotranspiration index,especially from January to May,and with cloud cover from March to May.These findings suggest that moisture availability does not limit radial growth in Acer hyrcanum and that the precipitation and water surplus signals may instead reflect the influence of cloud cover,which reduces sunlight availability during critical early-season months.This study contributes to the growing body of dendroclimatic research in the Alborz Mountains and,more broadly,on Acer species,particularly in high-elevation ecosystems where such studies are scarce.It also provides valuable insights into how Acer hyrcanum may respond to future climate change.
基金supported by the IITP(Institute for Information&Communications Technology Planning&Evaluation)under the ITRC(Information Technology Research Center)support program(IITP-2025-RS-2024-00438288)grant funded by the Korea government(MSIT)+1 种基金National Research Council of Science&Technology(NST)grant by the MSIT(Aerospace Semiconductor Strategy Research Project No.GTL25051-000)supported by the IC Design Education Center(IDEC),Korea。
文摘This work presents a systematic analysis of proton-induced total ionizing dose(TID)effects in 1.2 k V silicon carbide(SiC)power devices with various edge termination structures.Three edge terminations including ring-assisted junction termination extension(RA-JTE),multiple floating zone JTE(MFZ-JTE),and field limiting rings(FLR)were fabricated and irradiated with45 Me V protons at fluences ranging from 1×10^(12) to 1×10^(14) cm^(-2).Experimental results,supported by TCAD simulations,show that the RA-JTE structure maintained stable breakdown performance with less than 1%variation due to its effective electric field redistribution by multiple P+rings.In contrast,MFZ-JTE and FLR exhibit breakdown voltage shifts of 6.1%and 15.2%,respectively,under the highest fluence.These results demonstrate the superior radiation tolerance of the RA-JTE structure under TID conditions and provide practical design guidance for radiation-hardened Si C power devices in space and other highradiation environments.
基金supported by the University of Tabuk,Saudi Arabia。
文摘Multimodal spatiotemporal data from smart city consumer electronics present critical challenges including cross-modal temporal misalignment,unreliable data quality,limited joint modeling of spatial and temporal dependencies,and weak resilience to adversarial updates.To address these limitations,EdgeST-Fusion is introduced as a cross-modal federated graph transformer framework for context-aware smart city analytics.The architecture integrates cross-modal embedding networks for modality alignment,graph transformer encoders for spatial dependency modeling,temporal self-attention for dynamic pattern learning,and adaptive anomaly detection to ensure data quality and security during aggregation.A privacy-preserving federated learning protocol with differential privacy guarantees enables collaborative model training without centralizing sensitive data.The framework employs data-quality-aware weighted aggregation to enhance robustness against noisy and malicious client updates.Experimental evaluation on the GeoLife,PeMS-Bay,and SmartHome+datasets demonstrates that EdgeST-Fusion achieves 21.8%improvement in prediction accuracy,35.7%reduction in communication overhead,and 29.4%enhancement in security resilience compared to recent baselines.Real-world deployment across three smart city testbeds validates practical viability with 90.0%average accuracy and sub-250 ms inference latency.The proposed framework remains feasible for deployment on heterogeneous and resource-constrained consumer electronics devices whilemaintaining strong privacy guarantees and scalability for large-scale urban environments.
基金supported by the National Natural Science Foundation of China(Grant Nos.62572057,62272049,U24A20331)Beijing Natural Science Foundation(Grant Nos.4232026,4242020)Academic Research Projects of Beijing Union University(Grant No.ZK10202404).
文摘Traffic sign detection is a critical component of driving systems.Single-stage network-based traffic sign detection algorithms,renowned for their fast detection speeds and high accuracy,have become the dominant approach in current practices.However,in complex and dynamic traffic scenes,particularly with smaller traffic sign objects,challenges such as missed and false detections can lead to reduced overall detection accuracy.To address this issue,this paper proposes a detection algorithm that integrates edge and shape information.Recognizing that traffic signs have specific shapes and distinct edge contours,this paper introduces an edge feature extraction branch within the backbone network,enabling adaptive fusion with features of the same hierarchical level.Additionally,a shape prior convolution module is designed to replaces the first two convolutional modules of the backbone network,aimed at enhancing the model's perception ability for specific shape objects and reducing its sensitivity to background noise.The algorithm was evaluated on the CCTSDB and TT100k datasets,and compared to YOLOv8s,the mAP50 values increased by 3.0%and 10.4%,respectively,demonstrating the effectiveness of the proposed method in improving the accuracy of traffic sign detection.
基金supported by the Deanship of Graduate Studies and Scientific Research at Jouf University.
文摘Due to the growth of smart cities,many real-time systems have been developed to support smart cities using Internet of Things(IoT)and emerging technologies.They are formulated to collect the data for environment monitoring and automate the communication process.In recent decades,researchers have made many efforts to propose autonomous systems for manipulating network data and providing on-time responses in critical operations.However,the widespread use of IoT devices in resource-constrained applications and mobile sensor networks introduces significant research challenges for cybersecurity.These systems are vulnerable to a variety of cyberattacks,including unauthorized access,denial-of-service attacks,and data leakage,which compromise the network’s security.Additionally,uneven load balancing between mobile IoT devices,which frequently experience link interferences,compromises the trustworthiness of the system.This paper introduces a Multi-Agent secured framework using lightweight edge computing to enhance cybersecurity for sensor networks,aiming to leverage artificial intelligence for adaptive routing and multi-metric trust evaluation to achieve data privacy and mitigate potential threats.Moreover,it enhances the efficiency of distributed sensors for energy consumption through intelligent data analytics techniques,resulting in highly consistent and low-latency network communication.Using simulations,the proposed framework reveals its significant performance compared to state-of-the-art approaches for energy consumption by 43%,latency by 46%,network throughput by 51%,packet loss rate by 40%,and denial of service attacks by 42%.
文摘The Hall effect of elastic waves has attracted much attention due to its unique properties.A hexagonal lattice phononic crystal plate model is designed in this paper.By changing the spatial symmetry of the unit cell,a band gap for the A_(0) Lamb wave is opened.The existence of the edge state of the phononic crystal plate is obtained by finite element simulation.It is found that both zigzag-type edge and bridge edge are topological edge states by analysis of the band structure of the supercell.A rectangular model with a straight channel is designed and the simulation results show that the two types of channels are topologically protected only for the A_(0) mode Lamb wave but not for the S_(0) mode.In addition,the results of numerical simulation are verified by experimental data measured by a laser vibrometer.Finally,it is found that neither upside V-shaped channels nor channels with defects will affect the stable propagation of A_(0) Lamb waves along the proposed route.This proposed model and method are helpful in broadening the means of regulating elastic waves in phononic crystal structures,and extending practical application of topological edge states in such structures.
文摘With the rapid expansion of drone applications,accurate detection of objects in aerial imagery has become crucial for intelligent transportation,urban management,and emergency rescue missions.However,existing methods face numerous challenges in practical deployment,including scale variation handling,feature degradation,and complex backgrounds.To address these issues,we propose Edge-enhanced and Detail-Capturing You Only Look Once(EHDC-YOLO),a novel framework for object detection in Unmanned Aerial Vehicle(UAV)imagery.Based on the You Only Look Once version 11 nano(YOLOv11n)baseline,EHDC-YOLO systematically introduces several architectural enhancements:(1)a Multi-Scale Edge Enhancement(MSEE)module that leverages multi-scale pooling and edge information to enhance boundary feature extraction;(2)an Enhanced Feature Pyramid Network(EFPN)that integrates P2-level features with Cross Stage Partial(CSP)structures and OmniKernel convolutions for better fine-grained representation;and(3)Dynamic Head(DyHead)with multi-dimensional attention mechanisms for enhanced cross-scale modeling and perspective adaptability.Comprehensive experiments on the Vision meets Drones for Detection(VisDrone-DET)2019 dataset demonstrate that EHDC-YOLO achieves significant improvements,increasing mean Average Precision(mAP)@0.5 from 33.2%to 46.1%(an absolute improvement of 12.9 percentage points)and mAP@0.5:0.95 from 19.5%to 28.0%(an absolute improvement of 8.5 percentage points)compared with the YOLOv11n baseline,while maintaining a reasonable parameter count(2.81 M vs the baseline’s 2.58 M).Further ablation studies confirm the effectiveness of each proposed component,while visualization results highlight EHDC-YOLO’s superior performance in detecting objects and handling occlusions in complex drone scenarios.
基金supported by Institute of Information and Communications Technology Planning and Evaluation(IITP)grant funded by the Korean government(MSIT)(No.2019-0-01842,Artificial Intelligence Graduate School Program(GIST))supported by Korea Planning&Evaluation Institute of Industrial Technology(KEIT)grant funded by the Ministry of Trade,Industry&Energy(MOTIE,Republic of Korea)(RS-2025-25448249+1 种基金Automotive Industry Technology Development(R&D)Program)supported by the Regional Innovation System&Education(RISE)programthrough the(Gwangju RISE Center),funded by the Ministry of Education(MOE)and the Gwangju Metropolitan City,Republic of Korea(2025-RISE-05-001).
文摘In real-world autonomous driving tests,unexpected events such as pedestrians or wild animals suddenly entering the driving path can occur.Conducting actual test drives under various weather conditions may also lead to dangerous situations.Furthermore,autonomous vehicles may operate abnormally in bad weather due to limitations of their sensors and GPS.Driving simulators,which replicate driving conditions nearly identical to those in the real world,can drastically reduce the time and cost required for market entry validation;consequently,they have become widely used.In this paper,we design a virtual driving test environment capable of collecting and verifying SiLS data under adverse weather conditions using multi-source images.The proposed method generates a virtual testing environment that incorporates various events,including weather,time of day,and moving objects,that cannot be easily verified in real-world autonomous driving tests.By setting up scenario-based virtual environment events,multi-source image analysis and verification using real-world DCUs(Data Concentrator Units)with V2X-Car edge cloud can effectively address risk factors that may arise in real-world situations.We tested and validated the proposed method with scenarios employing V2X communication and multi-source image analysis.
基金the“Convergence and Open sharing System”Project,supported by the Ministry of Education and National Research Foundation of Korea.
文摘There has been an increasing emphasis on performing deep neural network(DNN)inference locally on edge devices due to challenges such as network congestion and security concerns.However,as DRAM process technology continues to scale down,the bit-flip errors in the memory of edge devices become more frequent,thereby leading to substantial DNN inference accuracy loss.Though several techniques have been proposed to alleviate the accuracy loss in edge environments,they require complex computations and additional parity bits for error correction,thus resulting in significant performance and storage overheads.In this paper,we propose FeatherGuard,a data-driven lightweight error protection scheme for DNN inference on edge devices.FeatherGuard selectively protects critical bit positions(that have a significant impact on DNN inference accuracy)against bit-flip errors,by considering various DNN characteristics(e.g.,data format,layer-wise weight distribution,actually stored logical values).Thus,it achieves high error tolerability during DNN inference.Since FeatherGuard reduces the bit-flip errors based on only a few simple arithmetic operations(e.g.,NOT operations)without parity bits,it causes negligible performance overhead and no storage overhead.Our experimental results show that FeatherGuard improves the error tolerability by up to 6667×and 4000×,compared to the conventional systems and the state-of-the-art error protection technique for edge environments,respectively.
文摘Industrial operators need reliable communication in high-noise,safety-critical environments where speech or touch input is often impractical.Existing gesture systems either miss real-time deadlines on resourceconstrained hardware or lose accuracy under occlusion,vibration,and lighting changes.We introduce Industrial EdgeSign,a dual-path framework that combines hardware-aware neural architecture search(NAS)with large multimodalmodel(LMM)guided semantics to deliver robust,low-latency gesture recognition on edge devices.The searched model uses a truncated ResNet50 front end,a dimensional-reduction network that preserves spatiotemporal structure for tubelet-based attention,and localized Transformer layers tuned for on-device inference.To reduce reliance on gloss annotations and mitigate domain shift,we distill semantics from factory-tuned vision-language models and pre-train with masked language modeling and video-text contrastive objectives,aligning visual features with a shared text space.OnML2HP and SHREC’17,theNAS-derived architecture attains 94.7% accuracywith 86ms inference latency and about 5.9W power on Jetson Nano.Under occlusion,lighting shifts,andmotion blur,accuracy remains above 82%.For safetycritical commands,the emergency-stop gesture achieves 72 ms 99th percentile latency with 99.7% fail-safe triggering.Ablation studies confirm the contribution of the spatiotemporal tubelet extractor and text-side pre-training,and we observe gains in translation quality(BLEU-422.33).These results show that Industrial EdgeSign provides accurate,resource-aware,and safety-aligned gesture recognition suitable for deployment in smart factory settings.
基金supported by the National Natural Science Foundation of China 62402171.
文摘As a fundamental component in computer vision,edges can be categorized into four types based on discontinuities in reflectance,illumination,surface normal,or depth.While deep CNNs have significantly advanced generic edge detection,real-time multi-class semantic edge detection under resource constraints remains challenging.To address this,we propose a lightweight framework based on PiDiNet that enables fine-grained semantic edge detection.Our model simultaneously predicts background and four edge categories from full-resolution inputs,balancing accuracy and efficiency.Key contributions include:a multi-channel output structure expanding binary edge prediction to five classes,supported by a deep supervision mechanism;a dynamic class-balancing strategy combining adaptive weighting with physical priors to handle extreme class imbalance;and maintained architectural efficiency enabling real-time inference.Extensive evaluations on BSDS-RIND show our approach achieves accuracy competitive with state-of-the-art methods while operating in real time.
基金supported by Jiangxi Polytechnic Institute Key Research Topics in Educational Reform 2025-JGJG-07.
文摘The intelligent environmental sensing systems are quickly transforming the sparse and retrospective monitoring to dense and decision-oriented environmental intelligence.This review brings together the manner in which integration of Internet of Things(IoT)sensing,edge computing,and real-time analytics facilitates timely detection,interpretation,and prediction of the environmental conditions across the applications,such as urban air quality,watershed and coastal surveillance,industrial safety,agriculture,and disaster response.We define end-to-end architectural patterns to organize devices,edge nodes,and cloud services to satisfy latency,reliability,bandwidth,and governance constraints with emphasis on event-time processing,adaptive offloading,and hierarchical aggregation.Then we look at sensing and infrastructure foundations,emphasizing the effects of sensor modality and power autonomy,connectivity,and the practices of calibration on the practicable analytics and eventual plausibility.It is on this basis that we examine real-time analytics pipelines and Artificial Intelligence(AI)techniques to preprocess,sensor combine,anomaly detect,and short-horizon forecast,with a focus on edge-deployable models,quantification of uncertainties,and query resistance to drift and domain shift.Lastly,we address the realities of deployment that condition operational success,such as lifecycle engineering,provenance-aware data management,security and privacy risks,ethical governance,and evaluation methodologies,which place end-to-end latency and field generalization as a priority.This review offers cohesion to algorithmic capabilities and systems engineering and governance to define an overall framework,show open areas of research directions,and provide practical recommendations on how to design trustworthy,scalable,and sustainable environmental monitoring systems.
基金supported in part by the National Natural Science Foundation of China(NSFC)under Grant 62276109The authors extend their appreciation to the Deanship of Scientific Research at King Saud University for funding this work through the Research Group Project number(ORF-2025-585).
文摘The personalized fine-tuning of large languagemodels(LLMs)on edge devices is severely constrained by limited computation resources.Although split federated learning alleviates on-device burdens,its effectiveness diminishes in few-shot reasoning scenarios due to the low data efficiency of conventional supervised fine-tuning,which leads to excessive communication overhead.To address this,we propose Language-Empowered Split Fine-Tuning(LESFT),a framework that integrates split architectures with a contrastive-inspired fine-tuning paradigm.LESFT simultaneously learns frommultiple logically equivalent but linguistically diverse reasoning chains,providing richer supervisory signals and improving data efficiency.This process-oriented training allows more effective reasoning adaptation with fewer samples.Extensive experiments demonstrate that LESFT consistently outperforms strong baselines such as SplitLoRA in task accuracy.LESFT consistently outperforms strong baselines on GSM8K,CommonsenseQA,and AQUA_RAT,with the largest gains observed on Qwen2.5-3B.These results indicate that LESFT can effectively adapt large language models for reasoning tasks under the computational and communication constraints of edge environments.
基金supported by the National Natural Science Foundation of China under Grant 62206083.
文摘Binary segmentation tasks in computer vision exhibit diverse appearance distributions and complex boundary characteristics.To address the limited generalization and adaptability of existing models across heterogeneous tasks,we propose Abel-Net,an Aggregated Bilateral Edge Localization Network designed as a universal framework formulti-task binary segmentation.Abel-Net integrates global and local contextual cues to enhance feature learning and edge precision.Specifically,amulti-scale feature pyramid fusion strategy is implemented via anAggregated Skip Connection(ASC)module to strengthen feature adaptability,while the Edge Dual Localization(EDL)mechanism performs coarse-to-fine refinement through edge-aware supervision.Additionally,Edge Attention(EA)and Edge Fusion Attention(EFA)modules prioritize edge-critical regions and facilitate accurate boundary alignment.Extensive experiments on nine diverse binary segmentation tasks demonstrate thatAbel-Net performs comparably to or surpasses state-of-the-art task-specific networks,exhibiting strong adaptability to a wide range of visual perception challenges.
基金supported by the Institute of Information&Communications Technology Planning&Evaluation(IITP)-Information Technology Research Center(ITRC)under Grant No.IITP-2025-RS-2023-00259004the Basic Science Research Program through the National Research Foundation of Korea(NRF)under Grant No.RS-2025-25434261.
文摘The intelligent transportation systems require secure,low-latency,and reliable communication architectures to enable the real-time vehicular application.This paper proposes an edge-intelligent semantic aggregation(EISA)framework for 6G unmanned aerial vehicle(UAV)-assisted Internet of vehicles(IoV)networks that integrates task-driven semantic communication,deep reinforcement learning(DRL)-based edge intelligence,and blockchain-based semantic validation across 6G terahertz(THz)links.UAVs in the proposed architecture serve as adaptive edge nodes that receive semantically vital information about the vehicle at any given stage,optimize aggregation and transmission parameters dynamically,and guarantee data integrity through a structured,lightweight consortium blockchain that signs semantically detailed representations rather than raw packets.Simulation results from a hybrid NS-3,MATLAB,and Python environment indicate that the proposed framework can achieve up to 45%reduction in end-to-end latency,an approximately 70%increase in throughput,and semantic efficiency with blockchain verification delays of less than 20 ms(more than 98%).These findings support the effectiveness of the proposed co-design for achieving context-aware,energy-efficient,and reliable communication under heavy-traffic conditions.The proposed framework provides a flexible and scalable foundation for next-generation 6G-enabled automotive networks,with subsequent growth toward federated learning-based collaborative intelligence,digital-twinassisted traffic modeling,and quantum-safe blockchain mechanisms to enhance scalability,intelligence,and long-term security.
