In response to the rising demand for low-latency,computation-intensive applications in vehicular networks,this paper proposes an adaptive task offloading approach for Vehicle-to-Everything(V2X)environments.Leveraging ...In response to the rising demand for low-latency,computation-intensive applications in vehicular networks,this paper proposes an adaptive task offloading approach for Vehicle-to-Everything(V2X)environments.Leveraging an enhanced Multi-Agent Deep Deterministic Policy Gradient(MADDPG)algorithm with an attention mechanism,the proposed approach optimizes computation offloading and resource allocation,aiming to minimize energy consumption and service delay.In this paper,vehicles dynamically offload computing-intensive tasks to both nearby vehicles through V2V links and roadside units through V2I links.The adaptive attention mechanism enables the system to prioritize relevant state information,leading to faster convergence.Simulations conducted in a realistic urban V2X scenario demonstrate that the proposed Attention-enhanced MADDPG(AT-MADDPG)algorithm significantly improves performance,achieving notable reductions in both energy consumption and latency compared to baseline algorithms,especially in high-demand,dynamic scenarios.展开更多
Large language model-based(LLM-based)text-to-SQL methods have achieved important progress in generating SQL queries for real-world applications.When confronted with table content-aware questions in real-world scenario...Large language model-based(LLM-based)text-to-SQL methods have achieved important progress in generating SQL queries for real-world applications.When confronted with table content-aware questions in real-world scenarios,ambiguous data content keywords and nonexistent database schema column names within the question lead to the poor performance of existing methods.To solve this problem,we propose a novel approach towards table content-aware text-to-SQL with self-retrieval(TCSR-SQL).It leverages LLM's in-context learning capability to extract data content keywords within the question and infer possible related database schema,which is used to generate Seed SQL to fuzz search databases.The search results are further used to confirm the encoding knowledge with the designed encoding knowledge table,including column names and exact stored content values used in the SQL.The encoding knowledge is sent to obtain the final Precise SQL following multirounds of generation-execution-revision process.To validate our approach,we introduce a table-content-aware,questionrelated benchmark dataset,containing 2115 question-SQL pairs.Comprehensive experiments conducted on this benchmark demonstrate the remarkable performance of TCSR-SQL,achieving an improvement of at least 27.8%in execution accuracy compared to other state-of-the-art methods.展开更多
Wide-field mesoscopy provides the capabilities of cortex-wide field of view(FOV),cellular resolution and high frame rate for neuronal imaging in the mouse brain.However,inherent background fluorescence degrades the im...Wide-field mesoscopy provides the capabilities of cortex-wide field of view(FOV),cellular resolution and high frame rate for neuronal imaging in the mouse brain.However,inherent background fluorescence degrades the image quality and hinders neuronal signal extraction.To address this problem,we first introduce a cortex-wide,high-resolution lineillumination mesoscope with a moving slit designed for in vivo mouse brain imaging.This system achieves a 6.6×6.6 mm FOV,microscale cellular resolution,a high frame rate of 10 Hz,as well as the background rejection ability.Furthermore,we integrated patterned illumination into the system to enhance the background suppression.Experimental results show that the proposed system successfully captures neurodynamics in the living mouse brain.Compared with conventional wide-field mesoscopes,the cortex-wide patterned line-illumination mesoscope(PLIM)achieves a threefold increase in the signal-to-background ratio(SBR).With patterned illumination integrated,the SBR enhancement further reaches four-anda-half-fold.展开更多
As an advanced device for observing atmospheric winds,the spaceborne Doppler Asymmetric Spatial Heterodyne(DASH)interferometer also encounters challenges associated with phase distortion,par-ticularly in limb sounding...As an advanced device for observing atmospheric winds,the spaceborne Doppler Asymmetric Spatial Heterodyne(DASH)interferometer also encounters challenges associated with phase distortion,par-ticularly in limb sounding scenarios.This paper discusses interferogram modeling and phase distortion cor-rection techniques for spaceborne DASH interferometers.The modeling of phase distortion interferograms with and without Doppler shift for limb observation was conducted,and the effectiveness of the analytical expression was verified through numerical simulation.The simulation results indicate that errors propagate layer by layer while using the onion-peeling inversion algorithm to handle phase-distorted interferograms.In contrast,the phase distortion correction algorithm can achieve effective correction.This phase correction method can be successfully applied to correct phase distortions in the interferograms of the spaceborne DASH interferometer,providing a feasible solution to enhance its measurement accuracy.展开更多
It is difficult to improve both energy consumption and detection accuracy simultaneously,and even to obtain the trade-off between them,when detecting and tracking moving targets,especially for Underwater Wireless Sens...It is difficult to improve both energy consumption and detection accuracy simultaneously,and even to obtain the trade-off between them,when detecting and tracking moving targets,especially for Underwater Wireless Sensor Networks(UWSNs).To this end,this paper investigates the relationship between the Degree of Target Change(DoTC)and the detection period,as well as the impact of individual nodes.A Hierarchical Detection and Tracking Approach(HDTA)is proposed.Firstly,the network detection period is determined according to DoTC,which reflects the variation of target motion.Secondly,during the network detection period,each detection node calculates its own node detection period based on the detection mutual information.Taking DoTC as pheromone,an ant colony algorithm is proposed to adaptively adjust the network detection period.The simulation results show that the proposed HDTA with the optimizations of network level and node level significantly improves the detection accuracy by 25%and the network energy consumption by 10%simultaneously,compared to the traditional adaptive period detection schemes.展开更多
Computational solid mechanics has become an indispensable approach in engineering,and numerical investigation of fracturing in composites is essential,as composites are widely used in structural applications.Crack evo...Computational solid mechanics has become an indispensable approach in engineering,and numerical investigation of fracturing in composites is essential,as composites are widely used in structural applications.Crack evolution in composites is the path to elucidating the relationship between microstructures and fracture performance,but crack-based finite-element methods are computationally expensive and time-consuming,which limits their application in computation-intensive scenarios.Consequently,this study proposes a deep learning framework called Crack-Net for instant prediction of the dynamic crack growth process,as well as its strain-stress curve.Specifically,Crack-Net introduces an implicit constraint technique,which incorporates the relationship between crack evolution and stress response into the network architecture.This technique substantially reduces data requirements while improving predictive accuracy.The transfer learning technique enables Crack-Net to handle composite materials with reinforcements of different strengths.Trained on high-accuracy fracture development datasets from phase field simulations,the proposed framework is capable of tackling intricate scenarios,involving materials with diverse interfaces,varying initial conditions,and the intricate elastoplastic fracture process.The proposed Crack-Net holds great promise for practical applications in engineering and materials science,in which accurate and efficient fracture prediction is crucial for optimizing material performance and microstructural design.展开更多
Advancements in mode-division multiplexing(MDM)techniques,aimed at surpassing the Shannon limit and augmenting transmission capacity,have garnered significant attention in optical fiber communica-tion,propelling the d...Advancements in mode-division multiplexing(MDM)techniques,aimed at surpassing the Shannon limit and augmenting transmission capacity,have garnered significant attention in optical fiber communica-tion,propelling the demand for high-quality multiplexers and demultiplexers.However,the criteria for ideal-mode multiplexers/demultiplexers,such as performance,scalability,compatibility,and ultra-compactness,have only partially been achieved using conventional bulky devices(e.g.,waveguides,grat-ings,and free space optics)—an issue that will substantially restrict the application of MDM techniques.Here,we present a neuro-meta-router(NMR)optimized through deep learning that achieves spatial multi-mode division and supports multi-channel communication,potentially offering scalability,com-patibility,and ultra-compactness.An MDM communication system based on an NMR is theoretically designed and experimentally demonstrated to enable simultaneous and independent multi-dataset transmission,showcasing a capacity of up to 100 gigabits per second(Gbps)and a symbol error rate down to the order of 104,all achieved without any compensation technologies or correlation devices.Our work presents a paradigm that merges metasurfaces,fiber communications,and deep learning,with potential applications in intelligent metasurface-aided optical interconnection,as well as all-optical pat-tern recognition and classification.展开更多
Dear Editor,This letter concentrates on distributed event-triggered formation control problems with finite-time convergence in an arbitrarily dimensional Euclidean space.A new unified approach of finite-time event-tri...Dear Editor,This letter concentrates on distributed event-triggered formation control problems with finite-time convergence in an arbitrarily dimensional Euclidean space.A new unified approach of finite-time event-triggered formation control is proposed by steering all agents to a sliding manifold(the affine image)to achieve general formations,like affine,rigid or translational formation.It only requires to design an extra steering law driving at least d+1 leaders from an affine image to a rigid or translational image,where d is the dimension of the space.The event-triggered function is designed in a distributed and discontinuous manner based only on local information to reduce the communication and calculation resources by aperiodic sampling.In the proposed event-triggered formation law,zeno-free behavior is ensured.展开更多
Purpose:This study investigates the impact of domestic mobility on Chinese scientists’academic performance and explores the predictors influencing their chances of moving to more prestigious institutions.Design/metho...Purpose:This study investigates the impact of domestic mobility on Chinese scientists’academic performance and explores the predictors influencing their chances of moving to more prestigious institutions.Design/methodology/approach:Using publication and affiliation data from OpenAlex,we identified 2,896 scientists who relocated between cities in China from 2014 to 2017.We applied propensity score matching(PSM)to compare their academic outcomes post-mobility with a matched group of non-mobile peers.Multiple performance metrics were examined,including publication count,citation impact,number of collaborators,and university prestige.Ordered logistic regression was used to analyze factors influencing moves to higher-level institutions.Findings:Mobility enhances collaboration by increasing the number of coauthors but is associated with a short-term decline in citation impact.Scientists were more likely to move to lower-prestige universities.However,prior collaboration breadth and citation count positively predicted transitions to more prestigious institutions,while the number of publications did not.Research limitations:This study focuses on intra-national mobility within China from 2014 to 2017 and relies on quantitative data,lacking personal or qualitative variables such as gender,discipline-specific norms,or institutional culture.Data coverage for Chinese-language publications may also be limited.Practical implications:This research provides insights into academic hiring patterns and the trade-offs involved in scientist mobility.It offers valuable guidance for institutions aiming to enhance faculty recruitment and retention,as well as for researchers considering career transitions.Originality/value:This is a quantitative analysis of domestic scientist mobility in China using matched comparison and multi-dimensional academic indicators.The integration of university prestige metrics(Double First-Class and citation-based rankings)offers a nuanced view of career dynamics within the Chinese higher education system.展开更多
Nonlinear transforms have significantly advanced learned image compression(LIC),particularly using residual blocks.This transform enhances the nonlinear expression ability and obtain compact feature representation by ...Nonlinear transforms have significantly advanced learned image compression(LIC),particularly using residual blocks.This transform enhances the nonlinear expression ability and obtain compact feature representation by enlarging the receptive field,which indicates how the convolution process extracts features in a high dimensional feature space.However,its functionality is restricted to the spatial dimension and network depth,limiting further improvements in network performance due to insufficient information interaction and representation.Crucially,the potential of high dimensional feature space in the channel dimension and the exploration of network width/resolution remain largely untapped.In this paper,we consider nonlinear transforms from the perspective of feature space,defining high-dimensional feature spaces in different dimensions and investigating the specific effects.Firstly,we introduce the dimension increasing and decreasing transforms in both channel and spatial dimensions to obtain high dimensional feature space and achieve better feature extraction.Secondly,we design a channel-spatial fusion residual transform(CSR),which incorporates multi-dimensional transforms for a more effective representation.Furthermore,we simplify the proposed fusion transform to obtain a slim architecture(CSR-sm),balancing network complexity and compression performance.Finally,we build the overall network with stacked CSR transforms to achieve better compression and reconstruction.Experimental results demonstrate that the proposed method can achieve superior ratedistortion performance compared to the existing LIC methods and traditional codecs.Specifically,our proposed method achieves 9.38%BD-rate reduction over VVC on Kodak dataset.展开更多
Complex industrial processes present typical uncertainty due to fluctuations in the composition of raw materials and frequently changing operating conditions.This poses three challenges for precise fault diagnosis,inc...Complex industrial processes present typical uncertainty due to fluctuations in the composition of raw materials and frequently changing operating conditions.This poses three challenges for precise fault diagnosis,including random noise interference,less distinguishability between multi-class faults,and the new fault emerging.To address these issues,this study formulates fault diagnosis in uncertain industrial processes as a multilevel refined fault diagnosis problem.A hierarchical stochastic network approach is proposed to refine fault diagnosis of multiclass faults.This method considers the augmentation of fault categories as naturally following a hierarchical structure.At each hierarchical stage,stochastic network methods are designed according to the sources of uncertainty.For fault feature extraction,a doubly stochastic attention-based variational graph autoencoder is introduced to suppress noise during the messagepassing process,ensuring the extraction of high-quality fault features and providing the provision of differentiated information.Subsequently,multiple stochastic configuration networks are deployed to realize multi-level fault diagnosis from coarse to fine granularity via a hierarchical structure rather than treating all faults equally.This approach effectively enhances the precision of multi-class fault diagnosis and ensures its robust generalization capability.Finally,the feasibility and effectiveness of the proposed method are validated using two industrial processes.The results demonstrate that the proposed method can effectively suppress the random noise interference and adapt to the emergence of small samples and imbalanced extreme fault-type data,achieving a satisfactory fault diagnosis performance.展开更多
The performance of traditional regular Intelligent Reflecting Surface(IRS)improves as the number of IRS elements increases,but more reflecting elements lead to higher IRS power consumption and greater overhead of chan...The performance of traditional regular Intelligent Reflecting Surface(IRS)improves as the number of IRS elements increases,but more reflecting elements lead to higher IRS power consumption and greater overhead of channel estimation.The Irregular Intelligent Reflecting Surface(IIRS)can enhance the performance of the IRS as well as boost the system performance when the number of reflecting elements is limited.However,due to the lack of radio frequency chain in IRS,it is challenging for the Base Station(BS)to gather perfect Channel State Information(CSI),especially in the presence of Eavesdroppers(Eves).Therefore,in this paper we investigate the minimum transmit power problem of IIRS-aided Simultaneous Wireless Information and Power Transfer(SWIPT)secure communication system with imperfect CSI of BS-IIRS-Eves links,which is subject to the rate outage probability constraints of the Eves,the minimum rate constraints of the Information Receivers(IRs),the energy harvesting constraints of the Energy Receivers(ERs),and the topology matrix constraints.Afterward,the formulated nonconvex problem can be efficiently tackled by employing joint optimization algorithm combined with successive refinement method and adaptive topology design method.Simulation results demonstrate the effectiveness of the proposed scheme and the superiority of IIRS.展开更多
High-fidelity tactile rendering offers significant potential for improving the richness and immersion of touchscreen interactions.This study focuses on a quantitative description of tactile rendering fidelity using a ...High-fidelity tactile rendering offers significant potential for improving the richness and immersion of touchscreen interactions.This study focuses on a quantitative description of tactile rendering fidelity using a custom-designed hybrid electrovibration and mechanical vibration(HEM)device.An electrovibration and mechanical vibration(EMV)algorithm that renders 3D gratings with different physical heights was proposed and shown to achieve 81%accuracy in shape recognition.Models of tactile rendering fidelity were established based on the evaluation of the height discrimination threshold,and the psychophysical-physical relationships between the discrimination and reference heights were well described by a modification of Weber’s law,with correlation coefficients higher than 0.9.The physiological-physical relationship between the pulse firing rate and the physical stimulation voltage was modeled using the Izhikevich spiking model with a logarithmic relationship.展开更多
As the 6G era approaches,wireless communication faces challenges such as massive user numbers,high mobility,and spectrum resource sharing.Radio maps are crucial for network design,optimization,and management,providing...As the 6G era approaches,wireless communication faces challenges such as massive user numbers,high mobility,and spectrum resource sharing.Radio maps are crucial for network design,optimization,and management,providing essential channel information.In this paper,we propose an innovative learning framework for Radio Map Estimation(RME)based on cycle-consistent generative adversarial networks.Traditional RME methods are often constrained by model complexity and interpolation accuracy,while learning-based methods require strictly paired datasets,making their practical application difficult.Our method overcomes these limitations by enabling training with unpaired data,efficiently converting local features into radio maps.Our experimental results demonstrate the effectiveness of the proposed method in two scenarios:accurate map data and map data with dynamic errors.To address dynamic interference,we designed a two-stage learning process that uses sparse observations to correct local details in the radio map,and the model's accuracy and practicality.展开更多
Real-world data always exhibit an imbalanced and long-tailed distribution,which leads to poor performance for neural network-based classification.Existing methods mainly tackle this problem by reweighting the loss fun...Real-world data always exhibit an imbalanced and long-tailed distribution,which leads to poor performance for neural network-based classification.Existing methods mainly tackle this problem by reweighting the loss function or rebalancing the classifier.However,one crucial aspect overlooked by previous research studies is the imbalanced feature space problem caused by the imbalanced angle distribution.In this paper,the authors shed light on the significance of the angle distribution in achieving a balanced feature space,which is essential for improving model performance under long-tailed distributions.Nevertheless,it is challenging to effectively balance both the classifier norms and angle distribution due to problems such as the low feature norm.To tackle these challenges,the authors first thoroughly analyse the classifier and feature space by decoupling the classification logits into three key components:classifier norm(i.e.the magnitude of the classifier vector),feature norm(i.e.the magnitude of the feature vector),and cosine similarity between the classifier vector and feature vector.In this way,the authors analyse the change of each component in the training process and reveal three critical problems that should be solved,that is,the imbalanced angle distribution,the lack of feature discrimination,and the low feature norm.Drawing from this analysis,the authors propose a novel loss function that incorporates hyperspherical uniformity,additive angular margin,and feature norm regularisation.Each component of the loss function addresses a specific problem and synergistically contributes to achieving a balanced classifier and feature space.The authors conduct extensive experiments on three popular benchmark datasets including CIFAR-10/100-LT,ImageNet-LT,and iNaturalist 2018.The experimental results demonstrate that the authors’loss function outperforms several previous state-of-the-art methods in addressing the challenges posed by imbalanced and longtailed datasets,that is,by improving upon the best-performing baselines on CIFAR-100-LT by 1.34,1.41,1.41 and 1.33,respectively.展开更多
Non-Terrestrial Networks(NTN)can be used to provide emergency voice services in Sixth-Generation(6G)communication systems.However,Internet of Things(Io T)terminals,which comprise restricted bandwidth resources and wea...Non-Terrestrial Networks(NTN)can be used to provide emergency voice services in Sixth-Generation(6G)communication systems.However,Internet of Things(Io T)terminals,which comprise restricted bandwidth resources and weak computing power,which make ensuring high-quality voice services over NTN challenging.Recent advancements in Artificial Intelligence(AI)techniques have been increasingly applied to enhance the audio quality and reduce the bit rate.However,applying models with high computational complexity to Io T terminals is difficult.In this study,we propose a voice-services-over NTN solution including a novel 6G non-terrestrial and ground network integrated framework and a lightweight Large Models(LMs)-driven codec operating at 450 bits per second.We also designed a new voice packet header and deployed an agent on-ground gateway to reduce the bandwidth overhead.The non-standard Session Initiation Protocol header was converted to the standard format while re-encapsulating Internet Protocol and User Datagram Protocol headers,replacing the conventional implementations.Additionally,an operational NTN satellite was used to evaluate the proposed Re Codec.The experimental results demonstrate that the Re Codec decreases the computational complexity by 96.61%while increasing the voice quality by 17.55%when compared with the state-of-the-art mechanisms.Furthermore,the design of the packet header reduced the voice frame header to 50 bytes.展开更多
Current research on reconfigurable parallel mechanisms(RPMs)primarily focuses on achieving limited configuration changes,while mechanisms capable of extensive mode switching with distinct motion branches remain challe...Current research on reconfigurable parallel mechanisms(RPMs)primarily focuses on achieving limited configuration changes,while mechanisms capable of extensive mode switching with distinct motion branches remain challenging to design.Conventional kinematotropic chains offer limited reconfigurability,underscoring the need for novel designs that enable broader operational adaptability.In this research,a novel diamond-like chain(DLC)with metamorphic units is proposed developed from generalized diamond kinematotropic chains.By altering the axes of the metamorphic units,the DLC realizes three distinct configurations,each corresponding to one of five motion branches characterized by bifurcation and metamorphic transitions.This DLC serves as the fundamental building block for constructing a reconfigurable hybrid limb.Using screw theory,the constraint properties of the limb in its five phases are analyzed and classified into three types:unconstrained limbs,limbs applying constraint forces,and limbs applying constraint couples.Based on this analysis,a RPM consisting of three reconfigurable limbs is developed.Its reconfigurability stems from the inherent bifurcation and metamorphic capabilities of the DLC-based limbs.This research introduces a RPM capable of controlled switching among ten distinct motion modes,with mobility ranging from three to six degrees of freedom.The proposed mechanism demonstrates high versatility and practical feasibility,offering a promising solution for applications requiring variable motion characteristics and adaptive performance.展开更多
With the rapid development of Intelligent Transportation Systems(ITS),many new applications for Intelligent Connected Vehicles(ICVs)have sprung up.In order to tackle the conflict between delay-sensitive applications a...With the rapid development of Intelligent Transportation Systems(ITS),many new applications for Intelligent Connected Vehicles(ICVs)have sprung up.In order to tackle the conflict between delay-sensitive applications and resource-constrained vehicles,computation offloading paradigm that transfers computation tasks from ICVs to edge computing nodes has received extensive attention.However,the dynamic network conditions caused by the mobility of vehicles and the unbalanced computing load of edge nodes make ITS face challenges.In this paper,we propose a heterogeneous Vehicular Edge Computing(VEC)architecture with Task Vehicles(TaVs),Service Vehicles(SeVs)and Roadside Units(RSUs),and propose a distributed algorithm,namely PG-MRL,which jointly optimizes offloading decision and resource allocation.In the first stage,the offloading decisions of TaVs are obtained through a potential game.In the second stage,a multi-agent Deep Deterministic Policy Gradient(DDPG),one of deep reinforcement learning algorithms,with centralized training and distributed execution is proposed to optimize the real-time transmission power and subchannel selection.The simulation results show that the proposed PG-MRL algorithm has significant improvements over baseline algorithms in terms of system delay.展开更多
To ensure the access security of 6G,physical-layer authentication(PLA)leverages the randomness and space-time-frequency uniqueness of the channel to provide unique identity signatures for transmitters.Furthermore,the ...To ensure the access security of 6G,physical-layer authentication(PLA)leverages the randomness and space-time-frequency uniqueness of the channel to provide unique identity signatures for transmitters.Furthermore,the introduction of artificial intelligence(AI)facilitates the learning of the distribution characteristics of channel fingerprints,effectively addressing the uncertainties and unknown dynamic challenges in wireless link modeling.This paper reviews representative AI-enabled PLA schemes and proposes a graph neural network(GNN)-based PLA approach in response to the challenges existing methods face in identifying mobile users.Simulation results demonstrate that the proposed method outperforms six baseline schemes in terms of authentication accuracy.Furthermore,this paper outlines the future development directions of PLA.展开更多
In this paper,we investigate the application of the Unmanned Aerial Vehicle(UAV)-enabled relaying system in emergency communications,where one UAV is applied as a relay to help transmit information from ground users t...In this paper,we investigate the application of the Unmanned Aerial Vehicle(UAV)-enabled relaying system in emergency communications,where one UAV is applied as a relay to help transmit information from ground users to a Base Station(BS).We maximize the total transmitted data from the users to the BS,by optimizing the user communication scheduling and association along with the power allocation and the trajectory of the UAV.To solve this non-convex optimization problem,we propose the traditional Convex Optimization(CO)and the Reinforcement Learning(RL)-based approaches.Specifically,we apply the block coordinate descent and successive convex approximation techniques in the CO approach,while applying the soft actor-critic algorithm in the RL approach.The simulation results show that both approaches can solve the proposed optimization problem and obtain good results.Moreover,the RL approach establishes emergency communications more rapidly than the CO approach once the training process has been completed.展开更多
基金supported by the National Key Research and Development Program of China under grant 2021YFA0716600。
文摘In response to the rising demand for low-latency,computation-intensive applications in vehicular networks,this paper proposes an adaptive task offloading approach for Vehicle-to-Everything(V2X)environments.Leveraging an enhanced Multi-Agent Deep Deterministic Policy Gradient(MADDPG)algorithm with an attention mechanism,the proposed approach optimizes computation offloading and resource allocation,aiming to minimize energy consumption and service delay.In this paper,vehicles dynamically offload computing-intensive tasks to both nearby vehicles through V2V links and roadside units through V2I links.The adaptive attention mechanism enables the system to prioritize relevant state information,leading to faster convergence.Simulations conducted in a realistic urban V2X scenario demonstrate that the proposed Attention-enhanced MADDPG(AT-MADDPG)algorithm significantly improves performance,achieving notable reductions in both energy consumption and latency compared to baseline algorithms,especially in high-demand,dynamic scenarios.
基金supported by the National Key Research and Development Program of China under(Grant 2023YFB3106504)Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies under(Grant 2022B1212010005)+2 种基金the Major Key Project of PCL under(Grant PCL2023A09)Shenzhen Science and Technology Program under(Grants ZDSYS20210623091809029 and RCBS20221008093131089)the project of Guangdong Power Grid Co.Ltd.under(Grants 037800KC23090005 and GD-KJXM20231042).
文摘Large language model-based(LLM-based)text-to-SQL methods have achieved important progress in generating SQL queries for real-world applications.When confronted with table content-aware questions in real-world scenarios,ambiguous data content keywords and nonexistent database schema column names within the question lead to the poor performance of existing methods.To solve this problem,we propose a novel approach towards table content-aware text-to-SQL with self-retrieval(TCSR-SQL).It leverages LLM's in-context learning capability to extract data content keywords within the question and infer possible related database schema,which is used to generate Seed SQL to fuzz search databases.The search results are further used to confirm the encoding knowledge with the designed encoding knowledge table,including column names and exact stored content values used in the SQL.The encoding knowledge is sent to obtain the final Precise SQL following multirounds of generation-execution-revision process.To validate our approach,we introduce a table-content-aware,questionrelated benchmark dataset,containing 2115 question-SQL pairs.Comprehensive experiments conducted on this benchmark demonstrate the remarkable performance of TCSR-SQL,achieving an improvement of at least 27.8%in execution accuracy compared to other state-of-the-art methods.
基金support from the National Natural Science Foundation of China(Grant No.61971256)。
文摘Wide-field mesoscopy provides the capabilities of cortex-wide field of view(FOV),cellular resolution and high frame rate for neuronal imaging in the mouse brain.However,inherent background fluorescence degrades the image quality and hinders neuronal signal extraction.To address this problem,we first introduce a cortex-wide,high-resolution lineillumination mesoscope with a moving slit designed for in vivo mouse brain imaging.This system achieves a 6.6×6.6 mm FOV,microscale cellular resolution,a high frame rate of 10 Hz,as well as the background rejection ability.Furthermore,we integrated patterned illumination into the system to enhance the background suppression.Experimental results show that the proposed system successfully captures neurodynamics in the living mouse brain.Compared with conventional wide-field mesoscopes,the cortex-wide patterned line-illumination mesoscope(PLIM)achieves a threefold increase in the signal-to-background ratio(SBR).With patterned illumination integrated,the SBR enhancement further reaches four-anda-half-fold.
文摘As an advanced device for observing atmospheric winds,the spaceborne Doppler Asymmetric Spatial Heterodyne(DASH)interferometer also encounters challenges associated with phase distortion,par-ticularly in limb sounding scenarios.This paper discusses interferogram modeling and phase distortion cor-rection techniques for spaceborne DASH interferometers.The modeling of phase distortion interferograms with and without Doppler shift for limb observation was conducted,and the effectiveness of the analytical expression was verified through numerical simulation.The simulation results indicate that errors propagate layer by layer while using the onion-peeling inversion algorithm to handle phase-distorted interferograms.In contrast,the phase distortion correction algorithm can achieve effective correction.This phase correction method can be successfully applied to correct phase distortions in the interferograms of the spaceborne DASH interferometer,providing a feasible solution to enhance its measurement accuracy.
文摘It is difficult to improve both energy consumption and detection accuracy simultaneously,and even to obtain the trade-off between them,when detecting and tracking moving targets,especially for Underwater Wireless Sensor Networks(UWSNs).To this end,this paper investigates the relationship between the Degree of Target Change(DoTC)and the detection period,as well as the impact of individual nodes.A Hierarchical Detection and Tracking Approach(HDTA)is proposed.Firstly,the network detection period is determined according to DoTC,which reflects the variation of target motion.Secondly,during the network detection period,each detection node calculates its own node detection period based on the detection mutual information.Taking DoTC as pheromone,an ant colony algorithm is proposed to adaptively adjust the network detection period.The simulation results show that the proposed HDTA with the optimizations of network level and node level significantly improves the detection accuracy by 25%and the network energy consumption by 10%simultaneously,compared to the traditional adaptive period detection schemes.
基金supported and partially funded by the National Natural Science Foundation of China(52288101)the China Postdoctoral Science Foundation(2024M761535)supported by the High Performance Computing Centers at Eastern Institute of Technology,Ningbo,and Ningbo Institute of Digital Twin.
文摘Computational solid mechanics has become an indispensable approach in engineering,and numerical investigation of fracturing in composites is essential,as composites are widely used in structural applications.Crack evolution in composites is the path to elucidating the relationship between microstructures and fracture performance,but crack-based finite-element methods are computationally expensive and time-consuming,which limits their application in computation-intensive scenarios.Consequently,this study proposes a deep learning framework called Crack-Net for instant prediction of the dynamic crack growth process,as well as its strain-stress curve.Specifically,Crack-Net introduces an implicit constraint technique,which incorporates the relationship between crack evolution and stress response into the network architecture.This technique substantially reduces data requirements while improving predictive accuracy.The transfer learning technique enables Crack-Net to handle composite materials with reinforcements of different strengths.Trained on high-accuracy fracture development datasets from phase field simulations,the proposed framework is capable of tackling intricate scenarios,involving materials with diverse interfaces,varying initial conditions,and the intricate elastoplastic fracture process.The proposed Crack-Net holds great promise for practical applications in engineering and materials science,in which accurate and efficient fracture prediction is crucial for optimizing material performance and microstructural design.
基金supported by the National Key Research and Development Program of China(2023YFB2804704)the National Natural Science Foundation of China(12174292,12374278,and 62105250).
文摘Advancements in mode-division multiplexing(MDM)techniques,aimed at surpassing the Shannon limit and augmenting transmission capacity,have garnered significant attention in optical fiber communica-tion,propelling the demand for high-quality multiplexers and demultiplexers.However,the criteria for ideal-mode multiplexers/demultiplexers,such as performance,scalability,compatibility,and ultra-compactness,have only partially been achieved using conventional bulky devices(e.g.,waveguides,grat-ings,and free space optics)—an issue that will substantially restrict the application of MDM techniques.Here,we present a neuro-meta-router(NMR)optimized through deep learning that achieves spatial multi-mode division and supports multi-channel communication,potentially offering scalability,com-patibility,and ultra-compactness.An MDM communication system based on an NMR is theoretically designed and experimentally demonstrated to enable simultaneous and independent multi-dataset transmission,showcasing a capacity of up to 100 gigabits per second(Gbps)and a symbol error rate down to the order of 104,all achieved without any compensation technologies or correlation devices.Our work presents a paradigm that merges metasurfaces,fiber communications,and deep learning,with potential applications in intelligent metasurface-aided optical interconnection,as well as all-optical pat-tern recognition and classification.
基金supported by the National Natural Science Foundation of China(62173118).
文摘Dear Editor,This letter concentrates on distributed event-triggered formation control problems with finite-time convergence in an arbitrarily dimensional Euclidean space.A new unified approach of finite-time event-triggered formation control is proposed by steering all agents to a sliding manifold(the affine image)to achieve general formations,like affine,rigid or translational formation.It only requires to design an extra steering law driving at least d+1 leaders from an affine image to a rigid or translational image,where d is the dimension of the space.The event-triggered function is designed in a distributed and discontinuous manner based only on local information to reduce the communication and calculation resources by aperiodic sampling.In the proposed event-triggered formation law,zeno-free behavior is ensured.
基金supported by grants from Shenzhen Polytechnic University Research(Fund No.6025310042 K)the National Natural Science Foundation of China(No.NSFC62006109 and NSFC12031005).
文摘Purpose:This study investigates the impact of domestic mobility on Chinese scientists’academic performance and explores the predictors influencing their chances of moving to more prestigious institutions.Design/methodology/approach:Using publication and affiliation data from OpenAlex,we identified 2,896 scientists who relocated between cities in China from 2014 to 2017.We applied propensity score matching(PSM)to compare their academic outcomes post-mobility with a matched group of non-mobile peers.Multiple performance metrics were examined,including publication count,citation impact,number of collaborators,and university prestige.Ordered logistic regression was used to analyze factors influencing moves to higher-level institutions.Findings:Mobility enhances collaboration by increasing the number of coauthors but is associated with a short-term decline in citation impact.Scientists were more likely to move to lower-prestige universities.However,prior collaboration breadth and citation count positively predicted transitions to more prestigious institutions,while the number of publications did not.Research limitations:This study focuses on intra-national mobility within China from 2014 to 2017 and relies on quantitative data,lacking personal or qualitative variables such as gender,discipline-specific norms,or institutional culture.Data coverage for Chinese-language publications may also be limited.Practical implications:This research provides insights into academic hiring patterns and the trade-offs involved in scientist mobility.It offers valuable guidance for institutions aiming to enhance faculty recruitment and retention,as well as for researchers considering career transitions.Originality/value:This is a quantitative analysis of domestic scientist mobility in China using matched comparison and multi-dimensional academic indicators.The integration of university prestige metrics(Double First-Class and citation-based rankings)offers a nuanced view of career dynamics within the Chinese higher education system.
基金supported by the Key Program of the National Natural Science Foundation of China(Grant No.62031013)Guangdong Province Key Construction Discipline Scientific Research Capacity Improvement Project(Grant No.2022ZDJS117).
文摘Nonlinear transforms have significantly advanced learned image compression(LIC),particularly using residual blocks.This transform enhances the nonlinear expression ability and obtain compact feature representation by enlarging the receptive field,which indicates how the convolution process extracts features in a high dimensional feature space.However,its functionality is restricted to the spatial dimension and network depth,limiting further improvements in network performance due to insufficient information interaction and representation.Crucially,the potential of high dimensional feature space in the channel dimension and the exploration of network width/resolution remain largely untapped.In this paper,we consider nonlinear transforms from the perspective of feature space,defining high-dimensional feature spaces in different dimensions and investigating the specific effects.Firstly,we introduce the dimension increasing and decreasing transforms in both channel and spatial dimensions to obtain high dimensional feature space and achieve better feature extraction.Secondly,we design a channel-spatial fusion residual transform(CSR),which incorporates multi-dimensional transforms for a more effective representation.Furthermore,we simplify the proposed fusion transform to obtain a slim architecture(CSR-sm),balancing network complexity and compression performance.Finally,we build the overall network with stacked CSR transforms to achieve better compression and reconstruction.Experimental results demonstrate that the proposed method can achieve superior ratedistortion performance compared to the existing LIC methods and traditional codecs.Specifically,our proposed method achieves 9.38%BD-rate reduction over VVC on Kodak dataset.
基金supported in part by the National Key Research and Development Program of China(2022YFB3304900)the Science and Technology Innovation Program of Hunan Province(2022RC1089)+1 种基金the Central South University Innovation Driven Research Programme(2023CXQD040)the Fundamental Research Funds for the Central Universities of Central South University(2025ZZTS0213).
文摘Complex industrial processes present typical uncertainty due to fluctuations in the composition of raw materials and frequently changing operating conditions.This poses three challenges for precise fault diagnosis,including random noise interference,less distinguishability between multi-class faults,and the new fault emerging.To address these issues,this study formulates fault diagnosis in uncertain industrial processes as a multilevel refined fault diagnosis problem.A hierarchical stochastic network approach is proposed to refine fault diagnosis of multiclass faults.This method considers the augmentation of fault categories as naturally following a hierarchical structure.At each hierarchical stage,stochastic network methods are designed according to the sources of uncertainty.For fault feature extraction,a doubly stochastic attention-based variational graph autoencoder is introduced to suppress noise during the messagepassing process,ensuring the extraction of high-quality fault features and providing the provision of differentiated information.Subsequently,multiple stochastic configuration networks are deployed to realize multi-level fault diagnosis from coarse to fine granularity via a hierarchical structure rather than treating all faults equally.This approach effectively enhances the precision of multi-class fault diagnosis and ensures its robust generalization capability.Finally,the feasibility and effectiveness of the proposed method are validated using two industrial processes.The results demonstrate that the proposed method can effectively suppress the random noise interference and adapt to the emergence of small samples and imbalanced extreme fault-type data,achieving a satisfactory fault diagnosis performance.
基金supported in part by the Shenzhen Basic Research Program under Grant JCYJ20220531103008018,and Grants 20231120142345001 and 20231127144045001the Natural Science Foundation of China under Grant U20A20156.
文摘The performance of traditional regular Intelligent Reflecting Surface(IRS)improves as the number of IRS elements increases,but more reflecting elements lead to higher IRS power consumption and greater overhead of channel estimation.The Irregular Intelligent Reflecting Surface(IIRS)can enhance the performance of the IRS as well as boost the system performance when the number of reflecting elements is limited.However,due to the lack of radio frequency chain in IRS,it is challenging for the Base Station(BS)to gather perfect Channel State Information(CSI),especially in the presence of Eavesdroppers(Eves).Therefore,in this paper we investigate the minimum transmit power problem of IIRS-aided Simultaneous Wireless Information and Power Transfer(SWIPT)secure communication system with imperfect CSI of BS-IIRS-Eves links,which is subject to the rate outage probability constraints of the Eves,the minimum rate constraints of the Information Receivers(IRs),the energy harvesting constraints of the Energy Receivers(ERs),and the topology matrix constraints.Afterward,the formulated nonconvex problem can be efficiently tackled by employing joint optimization algorithm combined with successive refinement method and adaptive topology design method.Simulation results demonstrate the effectiveness of the proposed scheme and the superiority of IIRS.
基金Supported by the National Natural Science Foundation of China under Grants 61631010 and 61806085.
文摘High-fidelity tactile rendering offers significant potential for improving the richness and immersion of touchscreen interactions.This study focuses on a quantitative description of tactile rendering fidelity using a custom-designed hybrid electrovibration and mechanical vibration(HEM)device.An electrovibration and mechanical vibration(EMV)algorithm that renders 3D gratings with different physical heights was proposed and shown to achieve 81%accuracy in shape recognition.Models of tactile rendering fidelity were established based on the evaluation of the height discrimination threshold,and the psychophysical-physical relationships between the discrimination and reference heights were well described by a modification of Weber’s law,with correlation coefficients higher than 0.9.The physiological-physical relationship between the pulse firing rate and the physical stimulation voltage was modeled using the Izhikevich spiking model with a logarithmic relationship.
基金supported in part by the Shenzhen Basic Research Program under Grant JCYJ20220531103008018,and Grants 20231120142345001 and 20231127144045001the Guangdong Basic Research Program under Grant 2024ZDZX1016。
文摘As the 6G era approaches,wireless communication faces challenges such as massive user numbers,high mobility,and spectrum resource sharing.Radio maps are crucial for network design,optimization,and management,providing essential channel information.In this paper,we propose an innovative learning framework for Radio Map Estimation(RME)based on cycle-consistent generative adversarial networks.Traditional RME methods are often constrained by model complexity and interpolation accuracy,while learning-based methods require strictly paired datasets,making their practical application difficult.Our method overcomes these limitations by enabling training with unpaired data,efficiently converting local features into radio maps.Our experimental results demonstrate the effectiveness of the proposed method in two scenarios:accurate map data and map data with dynamic errors.To address dynamic interference,we designed a two-stage learning process that uses sparse observations to correct local details in the radio map,and the model's accuracy and practicality.
基金National Key Research and Development Program of China,Grant/Award Numbers:2022YFB3103900,2023YFB3106504Major Key Project of PCL,Grant/Award Numbers:PCL2022A03,PCL2023A09+5 种基金Shenzhen Basic Research,Grant/Award Number:JCYJ20220531095214031Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies,Grant/Award Number:2022B1212010005Shenzhen International Science and Technology Cooperation Project,Grant/Award Number:GJHZ20220913143008015Natural Science Foundation of Guangdong Province,Grant/Award Number:2023A1515011959Shenzhen-Hong Kong Jointly Funded Project,Grant/Award Number:SGDX20230116091246007Shenzhen Science and Technology Program,Grant/Award Numbers:RCBS20221008093131089,ZDSYS20210623091809029。
文摘Real-world data always exhibit an imbalanced and long-tailed distribution,which leads to poor performance for neural network-based classification.Existing methods mainly tackle this problem by reweighting the loss function or rebalancing the classifier.However,one crucial aspect overlooked by previous research studies is the imbalanced feature space problem caused by the imbalanced angle distribution.In this paper,the authors shed light on the significance of the angle distribution in achieving a balanced feature space,which is essential for improving model performance under long-tailed distributions.Nevertheless,it is challenging to effectively balance both the classifier norms and angle distribution due to problems such as the low feature norm.To tackle these challenges,the authors first thoroughly analyse the classifier and feature space by decoupling the classification logits into three key components:classifier norm(i.e.the magnitude of the classifier vector),feature norm(i.e.the magnitude of the feature vector),and cosine similarity between the classifier vector and feature vector.In this way,the authors analyse the change of each component in the training process and reveal three critical problems that should be solved,that is,the imbalanced angle distribution,the lack of feature discrimination,and the low feature norm.Drawing from this analysis,the authors propose a novel loss function that incorporates hyperspherical uniformity,additive angular margin,and feature norm regularisation.Each component of the loss function addresses a specific problem and synergistically contributes to achieving a balanced classifier and feature space.The authors conduct extensive experiments on three popular benchmark datasets including CIFAR-10/100-LT,ImageNet-LT,and iNaturalist 2018.The experimental results demonstrate that the authors’loss function outperforms several previous state-of-the-art methods in addressing the challenges posed by imbalanced and longtailed datasets,that is,by improving upon the best-performing baselines on CIFAR-100-LT by 1.34,1.41,1.41 and 1.33,respectively.
基金supported in part by the Major Key Project of PCL under Grant PCL2023A07the Research and Development Program of China Telecom under Grant T-2025-27。
文摘Non-Terrestrial Networks(NTN)can be used to provide emergency voice services in Sixth-Generation(6G)communication systems.However,Internet of Things(Io T)terminals,which comprise restricted bandwidth resources and weak computing power,which make ensuring high-quality voice services over NTN challenging.Recent advancements in Artificial Intelligence(AI)techniques have been increasingly applied to enhance the audio quality and reduce the bit rate.However,applying models with high computational complexity to Io T terminals is difficult.In this study,we propose a voice-services-over NTN solution including a novel 6G non-terrestrial and ground network integrated framework and a lightweight Large Models(LMs)-driven codec operating at 450 bits per second.We also designed a new voice packet header and deployed an agent on-ground gateway to reduce the bandwidth overhead.The non-standard Session Initiation Protocol header was converted to the standard format while re-encapsulating Internet Protocol and User Datagram Protocol headers,replacing the conventional implementations.Additionally,an operational NTN satellite was used to evaluate the proposed Re Codec.The experimental results demonstrate that the Re Codec decreases the computational complexity by 96.61%while increasing the voice quality by 17.55%when compared with the state-of-the-art mechanisms.Furthermore,the design of the packet header reduced the voice frame header to 50 bytes.
基金Supported by National Natural Science Foundation of China(Grant No.52175019)Beijing Municipal Natural Science Foundation(Grant Nos.L222038,20240484699)+1 种基金Joint Funds of Industry-university-research of Shanghai Academy of Spaceflight Technology(Grant No.SAST2022-017)Beijing Municipal Key Laboratory of Space-ground Interconnection and Convergence of China and Key Laboratory of IoT Monitoring and Early Warning,Ministry of Emergency Management,Project‘Vice President of Science and Technology’of Changping District,Beijing.
文摘Current research on reconfigurable parallel mechanisms(RPMs)primarily focuses on achieving limited configuration changes,while mechanisms capable of extensive mode switching with distinct motion branches remain challenging to design.Conventional kinematotropic chains offer limited reconfigurability,underscoring the need for novel designs that enable broader operational adaptability.In this research,a novel diamond-like chain(DLC)with metamorphic units is proposed developed from generalized diamond kinematotropic chains.By altering the axes of the metamorphic units,the DLC realizes three distinct configurations,each corresponding to one of five motion branches characterized by bifurcation and metamorphic transitions.This DLC serves as the fundamental building block for constructing a reconfigurable hybrid limb.Using screw theory,the constraint properties of the limb in its five phases are analyzed and classified into three types:unconstrained limbs,limbs applying constraint forces,and limbs applying constraint couples.Based on this analysis,a RPM consisting of three reconfigurable limbs is developed.Its reconfigurability stems from the inherent bifurcation and metamorphic capabilities of the DLC-based limbs.This research introduces a RPM capable of controlled switching among ten distinct motion modes,with mobility ranging from three to six degrees of freedom.The proposed mechanism demonstrates high versatility and practical feasibility,offering a promising solution for applications requiring variable motion characteristics and adaptive performance.
基金supported by Future Network Scientific Research Fund Project (FNSRFP-2021-ZD-4)National Natural Science Foundation of China (No.61991404,61902182)+1 种基金National Key Research and Development Program of China under Grant 2020YFB1600104Key Research and Development Plan of Jiangsu Province under Grant BE2020084-2。
文摘With the rapid development of Intelligent Transportation Systems(ITS),many new applications for Intelligent Connected Vehicles(ICVs)have sprung up.In order to tackle the conflict between delay-sensitive applications and resource-constrained vehicles,computation offloading paradigm that transfers computation tasks from ICVs to edge computing nodes has received extensive attention.However,the dynamic network conditions caused by the mobility of vehicles and the unbalanced computing load of edge nodes make ITS face challenges.In this paper,we propose a heterogeneous Vehicular Edge Computing(VEC)architecture with Task Vehicles(TaVs),Service Vehicles(SeVs)and Roadside Units(RSUs),and propose a distributed algorithm,namely PG-MRL,which jointly optimizes offloading decision and resource allocation.In the first stage,the offloading decisions of TaVs are obtained through a potential game.In the second stage,a multi-agent Deep Deterministic Policy Gradient(DDPG),one of deep reinforcement learning algorithms,with centralized training and distributed execution is proposed to optimize the real-time transmission power and subchannel selection.The simulation results show that the proposed PG-MRL algorithm has significant improvements over baseline algorithms in terms of system delay.
文摘To ensure the access security of 6G,physical-layer authentication(PLA)leverages the randomness and space-time-frequency uniqueness of the channel to provide unique identity signatures for transmitters.Furthermore,the introduction of artificial intelligence(AI)facilitates the learning of the distribution characteristics of channel fingerprints,effectively addressing the uncertainties and unknown dynamic challenges in wireless link modeling.This paper reviews representative AI-enabled PLA schemes and proposes a graph neural network(GNN)-based PLA approach in response to the challenges existing methods face in identifying mobile users.Simulation results demonstrate that the proposed method outperforms six baseline schemes in terms of authentication accuracy.Furthermore,this paper outlines the future development directions of PLA.
基金supported in part by the Shenzhen Basic Research Project under Grant JCYJ20220531103008018 and Grant 20200812112423002in part by the Guangdong Basic Research Program under Grant 2019A1515110358,2021A1515012097in part by the open research fund of National Mobile Communications Research Laboratory,Southeast University (No.2021D16)。
文摘In this paper,we investigate the application of the Unmanned Aerial Vehicle(UAV)-enabled relaying system in emergency communications,where one UAV is applied as a relay to help transmit information from ground users to a Base Station(BS).We maximize the total transmitted data from the users to the BS,by optimizing the user communication scheduling and association along with the power allocation and the trajectory of the UAV.To solve this non-convex optimization problem,we propose the traditional Convex Optimization(CO)and the Reinforcement Learning(RL)-based approaches.Specifically,we apply the block coordinate descent and successive convex approximation techniques in the CO approach,while applying the soft actor-critic algorithm in the RL approach.The simulation results show that both approaches can solve the proposed optimization problem and obtain good results.Moreover,the RL approach establishes emergency communications more rapidly than the CO approach once the training process has been completed.
