This Special Topic of the Journal of Semiconductors(JoS)features expanded versions of key articles presented at the 2025 IEEE International Conference on Integrated Circuits Technologies and Applications(ICTA),which w...This Special Topic of the Journal of Semiconductors(JoS)features expanded versions of key articles presented at the 2025 IEEE International Conference on Integrated Circuits Technologies and Applications(ICTA),which was held in Macao,China,from October 22 to 24,2025.IEEE ICTA is an IEEE flagship conference in the field of integrated circuits(IC)in China,which provides a communication platform for sharing the state-of-the-art techniques from experts in the field of ICs.Among the 146 papers presented at ICTA 2025,the Technical Program Committee and the Award Committee have selected 3 high-quality articles for recommending to the Special Topic of JoS,covering the technical fields of RF,medical neural interface,and vision sensing ICs.展开更多
To effectively suppress the impact of laser phase noise on the extracted phase signal as required for distributed vibration sensing,we propose and experimentally demonstrate,for the first time,to our knowledge,a selfc...To effectively suppress the impact of laser phase noise on the extracted phase signal as required for distributed vibration sensing,we propose and experimentally demonstrate,for the first time,to our knowledge,a selfcoherent communication and sensing integrated system based on a balanced-delay architecture using a 7.6 MHz linewidth distributed feedback(DFB)laser.In the proposed scheme,identical delays are introduced at the transmitter and receiver to ensure precise optical path matching between the signal and the local oscillator(LO).This configuration effectively suppresses laser phase noise while retaining vibration-induced phase variations,thereby eliminating the need for costly ultra-narrow linewidth lasers.A 55.9 km weakly coupled seven-core fiber system is constructed,where the central core carries the LO and the six outer cores simultaneously transmit 14 Gbaud DP-16QAM signals,achieving multi-channel high-speed communication and distributed vibration sensing integration.The proposed system shares the same coherent transceiver,digital signal processing(DSP)modules,and spatial channels for both communication and sensing,enabling accurate detection of distributed vibration sensing with a frequency range of 900 Hz to 70 kHz.The system yields a phase responsivity of 0.235 rad/V at 10 kHz.This work provides a cost-effective and scalable approach toward highly integrated fiber networks for joint communication and sensing applications.展开更多
Cases of widespread bone hydatid infection are relatively rare in clinical practice.In this study,we reported for the first time a validated integrated repair therapy for multiple bone tissues,including the hip,femur,...Cases of widespread bone hydatid infection are relatively rare in clinical practice.In this study,we reported for the first time a validated integrated repair therapy for multiple bone tissues,including the hip,femur,and knee,caused by echinococ cosis.Artificial intelligence(AI)was used to develop a targeted surgical plan and to design a personalized prosthesis.Finite element analysis(FEA)was used to optimize the mechanical effectiveness of a customized integrated replacement prosthesis and to model stress distribution in the surrounding bone.Three-dimensional(3 D)printing was used to fabricate a customized prosthesis.With the assistance of AI,FEA,and 3 D printing technology,a personalized surgical plan and customized prosthesis were successfully constructed based on the patient’s disease.This approach achieved a successful therapeutic effect,demonstrating that AI-assisted personalized medicine holds great promise for the future.展开更多
Solar-driven interfacial evaporation presents a promising approach to address global freshwater scarcity.Current challenges in photothermal membrane design lie in achieving concurrent optimization of high solar absorp...Solar-driven interfacial evaporation presents a promising approach to address global freshwater scarcity.Current challenges in photothermal membrane design lie in achieving concurrent optimization of high solar absorption,low thermal conductivity,and water transport,where existing materials fail to establish effective“water-heat-salt”synergistic regulation at the evaporation interface.This study develops a seamlessly integrated Janus membrane through growing hydrophilic Cu2−xS nanostructure on a hydrophobic carbon cloth substrate with carbon black coating(CB/CC).By precisely engineering the submicron pore architecture within the Cu2−xS layer,we established a synergistic optimization mechanism for interfacial water transport,heat management,and salt rejection.The resulting Janus membrane demonstrates a high evaporation rate of 2.22 kg m−2h−1under 1 sun with an energy efficiency of about 88.4%.Notably,the system maintains stable operation in hypersaline environments(20 wt%NaCl)and achieves continuous 5-h salt-resistant evaporation.Moreover,the Janus membranes can effectively purify various industrial wastewater,including acidic,alkaline,and organic pollutants.This study provides a new strategy for developing high-efficiency portable desalination systems through interfacial engineering of pore architecture.展开更多
Background This study evaluated the impact of the Medicine IN Geriatric(MINE)application,which integrates the Integrated Medicine Management(IMM)model and STOPP/START criteria,on improving prescribing practices in eld...Background This study evaluated the impact of the Medicine IN Geriatric(MINE)application,which integrates the Integrated Medicine Management(IMM)model and STOPP/START criteria,on improving prescribing practices in elderly patients with congestive heart failure(CHF).Methods A two-phase study was conducted:validation of the IMM model and its implementation via the MINE app.A quasiexperimental pretest-posttest control group design was used in a hospital in East Kalimantan,Indonesia.Patients aged 60-79 years were randomly assigned to the intervention or control group.The intervention group received IMM-based pharmaceutical services,including medication reconciliation,repeated medication reviews,and individualized discharge counseling.The outcomes assessed included polypharmacy rates,potentially inappropriate medications(PIMs),potential prescribing omissions(PPOs),and health-related quality of life,using the EQ-5D-5L and EQ-VAS tools.Results Expert validation showed high content validity,with I-CVI≥0.86 and S-CVI=0.94.During implementation,the use of antiplatelets,statins,angiotensin converting enzyme inhibitors(ACEI),and angiotensin receptorβ-blockers(ARB)declined from admission to discharge.PIMs,such as beta-blockers in patients with conduction disorders and ACE-I/ARBs in those with hyperkalemia,also decreased.The intervention group's EQ-5D-5L scores improved from 0.552 to 0.664,whereas the control group's scores declined slightly.EQ-VAS scores also increased significantly in the intervention group.Conclusion The MINE-based IMM intervention effectively reduced inappropriate prescribing and enhanced the quality of life in elderly patients with CHF.This technology-enabled multidisciplinary approach supports safer prescribing in geriatric care.展开更多
The rapid expansion of the low-altitude economy is driving strong demand for highly accurate and reliable positioning technologies to support diverse aerial operations.This review examines core positioning methodologi...The rapid expansion of the low-altitude economy is driving strong demand for highly accurate and reliable positioning technologies to support diverse aerial operations.This review examines core positioning methodologies within the low-altitude intelligent network(LAIN)framework,beginning with an analysis of positioning requirements and performance metrics for low-altitude flight scenarios.It systematically assesses the principles,strengths,and limitations of mainstream positioning systems,including Global Navigation Satellite Systems(GNSS),terrestrial wireless positioning,and autonomous navigation,and it surveys prevalent integrated and cooperative positioning schemes.Our analysis demonstrates that standalone positioning technologies are inadequate in complex low-altitude settings,underscoring the pivotal role of multi-source fusion and unmanned aerial vehicle(UAV)swarm cooperative positioning as future trends.To address infrastructure gaps and high deployment costs in current LAIN systems,we propose a“space−air−ground”integrated and cooperative positioning architecture centered on GNSS and the 5th generation mobile communication technology(5G).The ground layer integrates 5G and GNSS for wide-area enhanced positioning.The aerial layer uses 5G aircraft-to-everything(A2X)and sidelink(SL)communications to build self-organizing networks for cooperative UAV localization.The space layer leverages low Earth orbit(LEO)satellites to overcome coverage limitations in communication and positioning.This hierarchical architecture reduces deployment costs through infrastructure reuse and enables deep integration of communication and navigation capabilities.By supporting collaborative enhancement across all three domains,the framework improves positioning robustness and delivers cost-effective,ubiquitous,and highly reliable positioning services.Finally,we outline promising research directions.This review aims to provide a systematic reference and a novel architectural perspective for the ongoing development of LAIN.展开更多
Integrated Internet of Things(IoT)brings novel opportunities for pervasive smart services,as these systems al-low for seamless information and resource sharing among IoT devices.Meanwhile,Federated Learning(FL)has eme...Integrated Internet of Things(IoT)brings novel opportunities for pervasive smart services,as these systems al-low for seamless information and resource sharing among IoT devices.Meanwhile,Federated Learning(FL)has emerged as a new framework for distributed deployment of machine learning models and a promising approach for implementing intelligent IoT systems.However,integrated IoT systems are usually composed of diverse IoT devices from different systems,and thus their ownership,roles,data distribution,and capabilities are heteroge-neous.Current FL algorithms mainly focus on handling non-Independent and Identically Distributed(Non-IID)issues,but often result in reduced and unsustainable performance in integrated IoT systems.Therefore,we inves-tigate in this paper the problem of personalized and sustainable FL in integrated IoT systems.First,we argue that different parties in integrated IoT are heterogeneous and limited in available resources for FL,and these parties are also selfish and expect rational outcomes during cooperation,which is essential for guaranteeing the sustain-ability of integrated IoT.Then,this paper provides a novel framework for device selection in FL.It first sets one instance of the model for each device,and iteratively selects devices to participate in model training based on the joint consideration of local model accuracy,similarity of parameters,and remaining resources per device.The proposed method guarantees the rational allocation of resources to maintain balanced model performance across all devices.In this way,the sustainability of the whole IoT system is improved such that no devices will suffer extreme resource exhaustion or poor performance.Finally,extensive evaluation is conducted to validate the advanced performance of the proposed method in integrated IoT systems.展开更多
To address the modeling fragmentation and predictive deviation caused by the conventional"singlemechanism,weakly coupled,additive response"approach in formation damage research,this study proposes an integra...To address the modeling fragmentation and predictive deviation caused by the conventional"singlemechanism,weakly coupled,additive response"approach in formation damage research,this study proposes an integrated modeling framework for multi-mechanism coupling throughout the entire drilling and completion process.Five dominant damage mechanisms are unified into a multi-physics formulation featuring a dual solid–liquid module architecture and a shared-state coupling mechanism.A structural-state integrated damage function(SSIDF)is introduced to establish a continuous mapping from microscopic mechanism evolution to macroscopic permeability degradation.A feedback network encompassing scaling,clay swelling,and water blocking is further developed,achieving bidirectional dynamic coupling among reaction kinetics,interfacial transport,and saturation fields,and representing one of the most systematic coupling schemes currently known.The model is solved via a space-time multi-scale optimization strategy,ensuring strong numerical stability and scalability.Field validation demonstrates a prediction accuracy of 98.6%,representing an improvement of over 8%compared to traditional additive models.The model is particularly applicable to unconventional reservoirs such as deepwater formations,where multi-mechanism damage evolves rapidly and conventional additive models fail to capture dynamic coupling behavior.展开更多
Continuous and highly accurate navigation of long-endurance flight vehicles continues to be a substantial challenge in Global Navigation Satellite Systems(GNSS)-denied environments.Though the integrated navigation bas...Continuous and highly accurate navigation of long-endurance flight vehicles continues to be a substantial challenge in Global Navigation Satellite Systems(GNSS)-denied environments.Though the integrated navigation based on multiple sensors is used to improve the navigation performance,the existing methods are prone to model mismatch and error accumulation under heterogeneous conditions of sensors.In this paper,a Tightly-coupled Interactive Multi-Model Factor Graph Optimization(TIMMFGO) navigation method is proposed to solve the problem.The developed integrated navigation framework consists of Inertial Navigation Systems(INS),Celestial Navigation Systems(CNS),Radio Navigation Systems(RNS),and Barometric Altimeters(BA).We propose a CNS/INS tightly-coupled graph architecture that integrates star vector observations with INS pre-integration,enabling dynamic compensation of gyroscopic bias while correcting the attitude update accuracy of INS.Then,an Interactive Multi-Model(IMM) adaptive weighting strategy is used to combine the vertical RNS factor with the BA factor for position,which can effectively reduce the altitude bias induced by the spatial configuration constraints of RNS.The simulation demonstrates that compared to the Huber M-estimation-based FGO(HMFGO),Windowing Anomaly-Detection-based FGO(WADFGO) and IMM Unscented Kalman Filter(IMMUKF)methods,the TIMMFGO method improves attitude accuracy by 46.49 %,25.68 % and 20.67 %,respectively,while correspondingly reducing position accuracy by 29.29 %,10.79 % and 6.96 %.展开更多
The commercial AM60(Mg−6Al−0.3Mn)die-casting alloy was modified through Mn,Ce,and La micro-alloying,each at a content below 0.2 wt.%.SEM,TEM,and Micro-CT were employed to characterize the microstructures and propertie...The commercial AM60(Mg−6Al−0.3Mn)die-casting alloy was modified through Mn,Ce,and La micro-alloying,each at a content below 0.2 wt.%.SEM,TEM,and Micro-CT were employed to characterize the microstructures and properties of AM60 based alloys.AM60-0.2La alloy showed excellent mechanical properties.The ultimate tensile strength,yield strength,and elongation of(288.0±1.7)MPa,(158.0±1.0)MPa,and(22.0±3.0)%were achieved in AM60-0.2La alloy.Besides,AM60-0.2La alloy exhibited the best corrosion resistance(0.29 mm/a)and fluidity among the investigated four alloys.The excellent mechanical properties and corrosion resistance are mainly attributed to the grain refinement strengthening,low porosity,and low content of large shrinkage porosity,promising for super-sized integrated automotive components.展开更多
Phytomelatonin,an emerging plant hormone,plays vital roles in plant growth,development,and stress adaptation(Arnao et al.,2022;Ullah et al.,2024).It acts both as a direct antioxidant and a signaling molecule,engaging ...Phytomelatonin,an emerging plant hormone,plays vital roles in plant growth,development,and stress adaptation(Arnao et al.,2022;Ullah et al.,2024).It acts both as a direct antioxidant and a signaling molecule,engaging complex networks and interacting with other phytohormones(Liu et al.,2022;Khan et al.,2023).Although phytomelatonin receptors(PMTRs)have been identified in many plants(Wei et al.,2018;Wang et al.,2022;Liu et al.,2025),the downstream signaling mechanisms,particularly receptor-mediated protein modifications and transcriptional regulation,remain poorly characterized.展开更多
1 Current status and challenges The airway is a structurally and functionally complex system comprising stem and differentiated epithelial cell types that together form the critical airway-blood barrier through intera...1 Current status and challenges The airway is a structurally and functionally complex system comprising stem and differentiated epithelial cell types that together form the critical airway-blood barrier through interactions between the epithelial and endothelial layers.This system is supported by a diverse array of stromal cells embedded in an extracellular matrix(ECM).展开更多
Carbon-based air cathodes offer low cost,high electrical conductivity,and structural tunability.However,they suffer from limited catalytic activity and inefficient gas transport,and they typically rely on noble metal ...Carbon-based air cathodes offer low cost,high electrical conductivity,and structural tunability.However,they suffer from limited catalytic activity and inefficient gas transport,and they typically rely on noble metal additives or complex multilayer configurations.To tackle these issues,this study devised a self-activated integrated carbon-based air cathode.By integrating in situ catalytic site construction with structural optimization,the strategy not only induces the formation of oxygen functional groups(─C─OH,─C═O,─COOH),hierarchical pores,and uniformly distributed active sites,but also establishes a favorable electronic and mass-transport environment.Furthermore,the roll-pressing-based integrated design streamlines electrode construction,reinforces interfacial bonding,and significantly enhances mechanical stability.Density functional theory(DFT)calculations show that oxygen functional groups initiate hydrogen bonding interaction and promote charge enrichment,which improves the activity of the cathode and facilitates intermediate adsorption/desorption in oxygen reduction and evolution reactions processes.As a result,the integrated air cathode-based rechargeable zinc-air batteries(RZABs)achieve a high specific capacity of 811 mAh g-1.It also performs well in quasi-solid-state RZABs and silicon-air batteries systems across a wide temperature range,demonstrating strong adaptability and application potential.This study provides a scalable and cost-effective design strategy for high-performance carbon-based air cathodes,offering new insights into advancing durable and practical metal-air energy systems.展开更多
Based on the demands for crashworthiness and lightweight in the passive safety of transportation vehicles,metal-fiber reinforced polymer(FRP)hybrid thin-walled tubes(MFHTWTs)integrate the toughness,strength and lightw...Based on the demands for crashworthiness and lightweight in the passive safety of transportation vehicles,metal-fiber reinforced polymer(FRP)hybrid thin-walled tubes(MFHTWTs)integrate the toughness,strength and lightweight of two distinct material characteristics.MFHTWTs can achieve energy absorption through the coupling of material plastic deformation and fracture,demonstrating significant engineering value in passive safety.This review provides a comprehensive examination of the crashworthiness topology optimization of MFHTWTs,aiming to demonstrate that a deeply integrated approach combining topology and parameter opti-mization can realize an optimal design method for MFHTWTs,thereby maximizing the functional utilization of limited material.Firstly,the review highlights the crashworthiness topology optimization methods(CTOMs)based on thin-walled structures.With a particular focus on metal,the review discusses both the practical ap-plicability and limitations of CTOMs under crash conditions.Additionally,based on the methodology of the equivalent static load method(ESLM),the review emphasizes that topology optimization methods considering continuous fiber paths and multi-material interface connections are also applicable to the crashworthiness op-timization of MFHTWTs.Furthermore,to couple structural parameters and configuration characteristics,in-tegrated topology optimization methods,including parameter optimization,are proposed to provide a valuable reference for the global optimization of MFHTWTs.Thus,these methods can establish the mapping relationship between key parameters and the structural energy absorption capacity.展开更多
Recently,the technique of Integrated Sensing and Communication(ISAC)has gained great attentions and is expected to enable more advanced applications in cellular Vehicle-to-Everything(V2X).We intend to use the periodic...Recently,the technique of Integrated Sensing and Communication(ISAC)has gained great attentions and is expected to enable more advanced applications in cellular Vehicle-to-Everything(V2X).We intend to use the periodical reference signals,such as the synchronization signals in 5G,to actively detect targets at different azimuths via beam scanning.This paper mainly aims to optimize the beamwidth of these reference signals in cellular-V2X to facilitate both sensing and communication functions.Firstly,to address the issue of insufficient beam duration for accurate estimation of Doppler and the corresponding velocity as well,we combine multi-ple beam scanning cycles to support improved sensing performance.In this context,an innovative beamwidth optimization algorithm is proposed.Specifically,we deduce the Fisher Information Matrix(FIM)of the suffi-cient statistic associated with target azimuths.Considering the randomness of the target position,we build the objective function based on the expected trace of the derived FIM.Additionally,we formulate the constraint of channel coherent time for efficient communications.It is worth noting that this is the first time to optimize the beamwidth of periodic reference signals for joint communication and sensing in cellular V2X.Experimental results show that the optimal beamwidth varies with the steering azimuth of antennas and reaches its maximum when the beam points at the direction perpendicular to the roadside.展开更多
Traditional dynamic analysis of mechanical structures,often limited to individual beams or plates,fails to fully capture their dynamic behaviors.In systems where space and mass are constrained,such as the battery supp...Traditional dynamic analysis of mechanical structures,often limited to individual beams or plates,fails to fully capture their dynamic behaviors.In systems where space and mass are constrained,such as the battery support structures in electric aircraft,conventional absorbers and isolators are insufficient for effective vibration control.This study simplifies the battery support structure of electric aircraft as an integrated composite beam consisting of three interconnected beams,and investigated its structural dynamics properties and nonlinear vibration control under thermal conditions caused by battery heat.The nonlinear vibration control is performed using the Nitinol steel wire ropes(Ni Ti-ST),with nonlinear damping properties.The natural frequencies of system are determined using the Rayleigh-Ritz technique.Theoretical results are validated through both Finite Element Method(FEM)and hammer tests.Moreover,the dynamic equations are derived using the Lagrange method and discretized via the Galerkin Truncation Method(GTM).The Harmonic Balance Method(HBM)is used to evaluate the vibration responses of the integrated model,with further verification through the Runge-Kutta Method(RKM).The experiments are conducted to corroborate the theoretical analysis.The results show that the system frequency changes in stages with the increase of the stiffness of the integrated composite beam connection.Especially in the case of varying environments,as the temperature increases,the frequency of system will first increase to a certain maximum value and then gradually decrease.Furthermore,the NiTi-ST effectively reduces vibration in the integrated composite beam,particularly under varying temperatures and external excitations.展开更多
Microbial Cr(Ⅵ)reduction represents a cost-effective remediation approach,yet its application in continuous-flow systems is often constrained by toxicity inhibition and nutrient instability.This study developed an in...Microbial Cr(Ⅵ)reduction represents a cost-effective remediation approach,yet its application in continuous-flow systems is often constrained by toxicity inhibition and nutrient instability.This study developed an integrated microbial—phosphorus mineral—corncob(MPC)composite,which encapsulates Cr(Ⅵ)-reducing bacteria with slow-release nutrients within a hydrogel matrix.The MPC system sustained complete Cr(Ⅵ)removal at influent concentrations up to 160 mg/L in synthetic wastewater and 120 mg/L in actual groundwater,conditions under which conventional systems typically fail.Reactive transport modeling revealed that diffusion limitations create protective intraparticle concentration gradients,while density functional theory(DFT)calculations confirmed strong Cr(Ⅲ)sequestration and sustained phosphorus bioavailability.The estimated treatment cost is competitive with existing technologies.These findings demonstrate that engineering a protected self-sufficient microbial microenvironment can overcome critical stability limitations in Cr(Ⅵ)bioremediation.展开更多
Silicon(Si)anodes offer an ultrahigh theoretical capacity but face two major barriers to commercialization:severe structural degradation caused by significant volume changes and sluggish ion transport kinetics due to ...Silicon(Si)anodes offer an ultrahigh theoretical capacity but face two major barriers to commercialization:severe structural degradation caused by significant volume changes and sluggish ion transport kinetics due to the discontinuous ionic conductance of conventional binders.To address these challenges,we develop a fully integrated Si anode using sulfonated graphene(SG)as a dual-functional ion-conductive and mechanical reinforcing framework within a conventional carboxymethyl cellulose/styrene-butadiene rubber(CMC/SBR)binder.Thermally activated reactions during electrode fabrication establish covalent sulfonate ester bonds between SG and CMC,elastic carboxylate crosslinks between SBR and CMC,and chemical anchoring between the binders and Si particles,which are all further reinforced by hydrogen bonding.This multi-bonding network not only dissipates mechanical stress and maintains electronic connectivity via embedded C65 carbon but also significantly enhances Li+transport through high intrinsic ionic conductivity of SG,facilitating the formation of a stable solid electrolyte interphase(SEI).The resulting Si@CMC/SBR/SG anode delivers a high initial capacity of 3513.2 mAh g-1and retains 77%capacity after 500 cycles at 2 A g-1.It achieves 762 mAh g-1at 4 A g-1and practical areal capacities exceeding 4 mAh cm-2.Full-cell tests with NCM811 cathodes confirm 86.2%capacity retention after 100 cycles.This work demonstrates a pragmatic and scalable integration paradigm for durable,high-energy-density Si anodes.展开更多
For more accessible and advanced health monitoring,the Body Area Network(BAN)design with semantic technologies offers efficient information sensing and communication in smart healthcare Artificial Intelligence of Thin...For more accessible and advanced health monitoring,the Body Area Network(BAN)design with semantic technologies offers efficient information sensing and communication in smart healthcare Artificial Intelligence of Things(AIoT).To address the critical challenges of effective communication and reduction of data transmission pressure in AIoT-BAN,a hybrid BAN system is proposed which enhances information processing and communication capabilities by leveraging semantic understanding and multimodal processing.It incorporates a semantic communication and sensing fusion framework,offloading based on the human Body Coupled Communication(BCC)channel,and multimodal semantic information integration to reduce data transmission pressure.The proposed method offers effective inclusive smart healthcare and daily health maintenance for the general public.展开更多
Transformer models face significant computational challenges in private inference(PI).Existing optimization methods often rely on isolated techniques,neglecting joint structural and operational improvements.We propose...Transformer models face significant computational challenges in private inference(PI).Existing optimization methods often rely on isolated techniques,neglecting joint structural and operational improvements.We propose IG-3D,a unified framework that integrates structured compression and operator approximation through accurate importance assessment.Our approach first evaluates attention head importance using Integrated Gradients(IG),offering greater stability and theoretical soundness than gradient-based methods.We then apply a threedimensional optimization:(1)structurally pruning redundant attention heads;(2)replacing Softmax with adaptive polynomial approximation to avoid exponential computations;(3)implementing layer-wise GELU substitution to accommodate different layer characteristics.A joint thresholdmechanism coordinates compression across dimensions under accuracy constraints.Experimental results on the GLUE benchmark show that our method achieves an average 2.9×speedup in inference latency and a 50%reduction in communication cost,while controlling the accuracy loss within 2.3%,demonstrating significant synergistic effects and a superior accuracy-efficiency trade-off compared to single-technique optimization strategies.展开更多
摘要This Special Topic of the Journal of Semiconductors(JoS)features expanded versions of key articles presented at the 2025 IEEE International Conference on Integrated Circuits Technologies and Applications(ICTA),which was held in Macao,China,from October 22 to 24,2025.IEEE ICTA is an IEEE flagship conference in the field of integrated circuits(IC)in China,which provides a communication platform for sharing the state-of-the-art techniques from experts in the field of ICs.Among the 146 papers presented at ICTA 2025,the Technical Program Committee and the Award Committee have selected 3 high-quality articles for recommending to the Special Topic of JoS,covering the technical fields of RF,medical neural interface,and vision sensing ICs.
基金National Natural Science Foundation of China(62201308)Pengcheng Laboratory+1 种基金Shenzhen Science and Technology Innovation Commission(RCBS20231211090611017,JCYJ20240813104835048)Guangdong Basic and Applied Basic Research Foundation(2025A1515010277)。
摘要To effectively suppress the impact of laser phase noise on the extracted phase signal as required for distributed vibration sensing,we propose and experimentally demonstrate,for the first time,to our knowledge,a selfcoherent communication and sensing integrated system based on a balanced-delay architecture using a 7.6 MHz linewidth distributed feedback(DFB)laser.In the proposed scheme,identical delays are introduced at the transmitter and receiver to ensure precise optical path matching between the signal and the local oscillator(LO).This configuration effectively suppresses laser phase noise while retaining vibration-induced phase variations,thereby eliminating the need for costly ultra-narrow linewidth lasers.A 55.9 km weakly coupled seven-core fiber system is constructed,where the central core carries the LO and the six outer cores simultaneously transmit 14 Gbaud DP-16QAM signals,achieving multi-channel high-speed communication and distributed vibration sensing integration.The proposed system shares the same coherent transceiver,digital signal processing(DSP)modules,and spatial channels for both communication and sensing,enabling accurate detection of distributed vibration sensing with a frequency range of 900 Hz to 70 kHz.The system yields a phase responsivity of 0.235 rad/V at 10 kHz.This work provides a cost-effective and scalable approach toward highly integrated fiber networks for joint communication and sensing applications.
基金partially supported by the National Natural Science Foundation of China(Nos.32471474 and 82102574)the Precision Medicine Project of People’s Hospital of Xinjiang Uygur Autonomous Region(No.20220305)+4 种基金Chengdu Advanced Metal Materials Industry Technology Research Institute Co.,Ltd.Support Project(No.24H0802)Sichuan Science and Technology Program(Nos.2025YFHZ0086,2023YFS0053,2024YFHZ0125,and 2025ZNSFSC0381)Project of Tianfu Jincheng Laboratory(No.2025ZH009)Guangdong Basic and Applied Basic Research Foundation(No.2023A1515220102)Xinjiang Autonomous Region Science and Technology Support Project Plan(Directive)Project(No.2024E02049)。
摘要Cases of widespread bone hydatid infection are relatively rare in clinical practice.In this study,we reported for the first time a validated integrated repair therapy for multiple bone tissues,including the hip,femur,and knee,caused by echinococ cosis.Artificial intelligence(AI)was used to develop a targeted surgical plan and to design a personalized prosthesis.Finite element analysis(FEA)was used to optimize the mechanical effectiveness of a customized integrated replacement prosthesis and to model stress distribution in the surrounding bone.Three-dimensional(3 D)printing was used to fabricate a customized prosthesis.With the assistance of AI,FEA,and 3 D printing technology,a personalized surgical plan and customized prosthesis were successfully constructed based on the patient’s disease.This approach achieved a successful therapeutic effect,demonstrating that AI-assisted personalized medicine holds great promise for the future.
基金the financial support of the National Natural Science Foundation of China(No.52075309)the Youth Innovation Team of Shaanxi Universities(21JP021).
摘要Solar-driven interfacial evaporation presents a promising approach to address global freshwater scarcity.Current challenges in photothermal membrane design lie in achieving concurrent optimization of high solar absorption,low thermal conductivity,and water transport,where existing materials fail to establish effective“water-heat-salt”synergistic regulation at the evaporation interface.This study develops a seamlessly integrated Janus membrane through growing hydrophilic Cu2−xS nanostructure on a hydrophobic carbon cloth substrate with carbon black coating(CB/CC).By precisely engineering the submicron pore architecture within the Cu2−xS layer,we established a synergistic optimization mechanism for interfacial water transport,heat management,and salt rejection.The resulting Janus membrane demonstrates a high evaporation rate of 2.22 kg m−2h−1under 1 sun with an energy efficiency of about 88.4%.Notably,the system maintains stable operation in hypersaline environments(20 wt%NaCl)and achieves continuous 5-h salt-resistant evaporation.Moreover,the Janus membranes can effectively purify various industrial wastewater,including acidic,alkaline,and organic pollutants.This study provides a new strategy for developing high-efficiency portable desalination systems through interfacial engineering of pore architecture.
基金supported by DRPM PDD 2023-2024,Ministry of Research,Technology and Higher Education of Indonesia(No.040/E5/PG.02.00.PL/2024,1675/B/UN3.LPPM/PT.01.03/2024)。
摘要Background This study evaluated the impact of the Medicine IN Geriatric(MINE)application,which integrates the Integrated Medicine Management(IMM)model and STOPP/START criteria,on improving prescribing practices in elderly patients with congestive heart failure(CHF).Methods A two-phase study was conducted:validation of the IMM model and its implementation via the MINE app.A quasiexperimental pretest-posttest control group design was used in a hospital in East Kalimantan,Indonesia.Patients aged 60-79 years were randomly assigned to the intervention or control group.The intervention group received IMM-based pharmaceutical services,including medication reconciliation,repeated medication reviews,and individualized discharge counseling.The outcomes assessed included polypharmacy rates,potentially inappropriate medications(PIMs),potential prescribing omissions(PPOs),and health-related quality of life,using the EQ-5D-5L and EQ-VAS tools.Results Expert validation showed high content validity,with I-CVI≥0.86 and S-CVI=0.94.During implementation,the use of antiplatelets,statins,angiotensin converting enzyme inhibitors(ACEI),and angiotensin receptorβ-blockers(ARB)declined from admission to discharge.PIMs,such as beta-blockers in patients with conduction disorders and ACE-I/ARBs in those with hyperkalemia,also decreased.The intervention group's EQ-5D-5L scores improved from 0.552 to 0.664,whereas the control group's scores declined slightly.EQ-VAS scores also increased significantly in the intervention group.Conclusion The MINE-based IMM intervention effectively reduced inappropriate prescribing and enhanced the quality of life in elderly patients with CHF.This technology-enabled multidisciplinary approach supports safer prescribing in geriatric care.
基金supported by the National Key Research&Development Program of China(No.2024YFB3910102).
摘要The rapid expansion of the low-altitude economy is driving strong demand for highly accurate and reliable positioning technologies to support diverse aerial operations.This review examines core positioning methodologies within the low-altitude intelligent network(LAIN)framework,beginning with an analysis of positioning requirements and performance metrics for low-altitude flight scenarios.It systematically assesses the principles,strengths,and limitations of mainstream positioning systems,including Global Navigation Satellite Systems(GNSS),terrestrial wireless positioning,and autonomous navigation,and it surveys prevalent integrated and cooperative positioning schemes.Our analysis demonstrates that standalone positioning technologies are inadequate in complex low-altitude settings,underscoring the pivotal role of multi-source fusion and unmanned aerial vehicle(UAV)swarm cooperative positioning as future trends.To address infrastructure gaps and high deployment costs in current LAIN systems,we propose a“space−air−ground”integrated and cooperative positioning architecture centered on GNSS and the 5th generation mobile communication technology(5G).The ground layer integrates 5G and GNSS for wide-area enhanced positioning.The aerial layer uses 5G aircraft-to-everything(A2X)and sidelink(SL)communications to build self-organizing networks for cooperative UAV localization.The space layer leverages low Earth orbit(LEO)satellites to overcome coverage limitations in communication and positioning.This hierarchical architecture reduces deployment costs through infrastructure reuse and enables deep integration of communication and navigation capabilities.By supporting collaborative enhancement across all three domains,the framework improves positioning robustness and delivers cost-effective,ubiquitous,and highly reliable positioning services.Finally,we outline promising research directions.This review aims to provide a systematic reference and a novel architectural perspective for the ongoing development of LAIN.
基金partly supported by National Natural Science Foundation of China under grant No.62372085,61802050Xinjiang Science and Technology Program(No.2022D01B185)the China Postdoctoral Science Foundation under Grant Number 2025M781461.
摘要Integrated Internet of Things(IoT)brings novel opportunities for pervasive smart services,as these systems al-low for seamless information and resource sharing among IoT devices.Meanwhile,Federated Learning(FL)has emerged as a new framework for distributed deployment of machine learning models and a promising approach for implementing intelligent IoT systems.However,integrated IoT systems are usually composed of diverse IoT devices from different systems,and thus their ownership,roles,data distribution,and capabilities are heteroge-neous.Current FL algorithms mainly focus on handling non-Independent and Identically Distributed(Non-IID)issues,but often result in reduced and unsustainable performance in integrated IoT systems.Therefore,we inves-tigate in this paper the problem of personalized and sustainable FL in integrated IoT systems.First,we argue that different parties in integrated IoT are heterogeneous and limited in available resources for FL,and these parties are also selfish and expect rational outcomes during cooperation,which is essential for guaranteeing the sustain-ability of integrated IoT.Then,this paper provides a novel framework for device selection in FL.It first sets one instance of the model for each device,and iteratively selects devices to participate in model training based on the joint consideration of local model accuracy,similarity of parameters,and remaining resources per device.The proposed method guarantees the rational allocation of resources to maintain balanced model performance across all devices.In this way,the sustainability of the whole IoT system is improved such that no devices will suffer extreme resource exhaustion or poor performance.Finally,extensive evaluation is conducted to validate the advanced performance of the proposed method in integrated IoT systems.
基金financially supported by National Natural Science Foundation of China(No.U23B2082)Oil&Gas Major Project(No.2025ZD1404600)supported by the China Scholarship Council(202406440017)for one year research at the University of Dundee。
摘要To address the modeling fragmentation and predictive deviation caused by the conventional"singlemechanism,weakly coupled,additive response"approach in formation damage research,this study proposes an integrated modeling framework for multi-mechanism coupling throughout the entire drilling and completion process.Five dominant damage mechanisms are unified into a multi-physics formulation featuring a dual solid–liquid module architecture and a shared-state coupling mechanism.A structural-state integrated damage function(SSIDF)is introduced to establish a continuous mapping from microscopic mechanism evolution to macroscopic permeability degradation.A feedback network encompassing scaling,clay swelling,and water blocking is further developed,achieving bidirectional dynamic coupling among reaction kinetics,interfacial transport,and saturation fields,and representing one of the most systematic coupling schemes currently known.The model is solved via a space-time multi-scale optimization strategy,ensuring strong numerical stability and scalability.Field validation demonstrates a prediction accuracy of 98.6%,representing an improvement of over 8%compared to traditional additive models.The model is particularly applicable to unconventional reservoirs such as deepwater formations,where multi-mechanism damage evolves rapidly and conventional additive models fail to capture dynamic coupling behavior.
基金co-supported by the Open Fund of National Natural Science Foundation of China(No.62401042)the Young Elite Scientists Sponsorship Program by China Association for Science and Technology(No.2023QNRC001)。
摘要Continuous and highly accurate navigation of long-endurance flight vehicles continues to be a substantial challenge in Global Navigation Satellite Systems(GNSS)-denied environments.Though the integrated navigation based on multiple sensors is used to improve the navigation performance,the existing methods are prone to model mismatch and error accumulation under heterogeneous conditions of sensors.In this paper,a Tightly-coupled Interactive Multi-Model Factor Graph Optimization(TIMMFGO) navigation method is proposed to solve the problem.The developed integrated navigation framework consists of Inertial Navigation Systems(INS),Celestial Navigation Systems(CNS),Radio Navigation Systems(RNS),and Barometric Altimeters(BA).We propose a CNS/INS tightly-coupled graph architecture that integrates star vector observations with INS pre-integration,enabling dynamic compensation of gyroscopic bias while correcting the attitude update accuracy of INS.Then,an Interactive Multi-Model(IMM) adaptive weighting strategy is used to combine the vertical RNS factor with the BA factor for position,which can effectively reduce the altitude bias induced by the spatial configuration constraints of RNS.The simulation demonstrates that compared to the Huber M-estimation-based FGO(HMFGO),Windowing Anomaly-Detection-based FGO(WADFGO) and IMM Unscented Kalman Filter(IMMUKF)methods,the TIMMFGO method improves attitude accuracy by 46.49 %,25.68 % and 20.67 %,respectively,while correspondingly reducing position accuracy by 29.29 %,10.79 % and 6.96 %.
基金financially supported by the National Key Research and Development Program of China(Nos.2022YFB3709300,2021YFB3701000)the National Natural Science Foundation of China(Nos.52271090,52071036,U2037601,U21A2048)+1 种基金Chongqing Science and Technology Commission,China(Nos.CSTB2022TIAD-KPX0021,CSTC2024YCJHBGZXM0164,CSTB2024TIAD-KPX0001)the Fundamental Research Funds for the Central Universities,China(No.2022CDJDX-002)。
摘要The commercial AM60(Mg−6Al−0.3Mn)die-casting alloy was modified through Mn,Ce,and La micro-alloying,each at a content below 0.2 wt.%.SEM,TEM,and Micro-CT were employed to characterize the microstructures and properties of AM60 based alloys.AM60-0.2La alloy showed excellent mechanical properties.The ultimate tensile strength,yield strength,and elongation of(288.0±1.7)MPa,(158.0±1.0)MPa,and(22.0±3.0)%were achieved in AM60-0.2La alloy.Besides,AM60-0.2La alloy exhibited the best corrosion resistance(0.29 mm/a)and fluidity among the investigated four alloys.The excellent mechanical properties and corrosion resistance are mainly attributed to the grain refinement strengthening,low porosity,and low content of large shrinkage porosity,promising for super-sized integrated automotive components.
基金supported by the grants from the Key Research and Development Program of Xinjiang Uygur autonomous region in China(Grant No.2023B02017)the National Key Research and Development Program of China(Grant No.2024YFD2300703)+1 种基金the financial support from the Beijing Rural Revitalization Agricultural Science and Technology Project(Grant No.NY2401080000),BAIC01-2025the 2115 Talent Development Program of China Agricultural University.
摘要Phytomelatonin,an emerging plant hormone,plays vital roles in plant growth,development,and stress adaptation(Arnao et al.,2022;Ullah et al.,2024).It acts both as a direct antioxidant and a signaling molecule,engaging complex networks and interacting with other phytohormones(Liu et al.,2022;Khan et al.,2023).Although phytomelatonin receptors(PMTRs)have been identified in many plants(Wei et al.,2018;Wang et al.,2022;Liu et al.,2025),the downstream signaling mechanisms,particularly receptor-mediated protein modifications and transcriptional regulation,remain poorly characterized.
基金supported by the Innovations Biotechnology and Biological Sciences Research Council(No.BB/W014564/1)supported by the NIHR Cambridge Biomedical Research Centre(No.NIHR203312).
摘要1 Current status and challenges The airway is a structurally and functionally complex system comprising stem and differentiated epithelial cell types that together form the critical airway-blood barrier through interactions between the epithelial and endothelial layers.This system is supported by a diverse array of stromal cells embedded in an extracellular matrix(ECM).
基金funded by the National Nature Science Foundation of China(62264006,62574102)“Thousand Talents Program”of Yunnan Province for Young Talents,Innovative Research Teams(in Science and Technology)in the University of Yunnan Province(IRTSTYN),XingDian Talent Support Program for Young Talents,and Frontier Research Team of Kunming University 2023,The Basic Research Project of Yunnan Province(Nos.202201AU070022)+2 种基金Kunming University Talent Introduction Fund(Nos.YJL20024)Yunnan Province Education Department Scientific Research Fund Project(Nos.2024Y759)Undergraduate Innovation and Entrepreneurship Training Program Project of Yunnan Provincial(202411393005)。
摘要Carbon-based air cathodes offer low cost,high electrical conductivity,and structural tunability.However,they suffer from limited catalytic activity and inefficient gas transport,and they typically rely on noble metal additives or complex multilayer configurations.To tackle these issues,this study devised a self-activated integrated carbon-based air cathode.By integrating in situ catalytic site construction with structural optimization,the strategy not only induces the formation of oxygen functional groups(─C─OH,─C═O,─COOH),hierarchical pores,and uniformly distributed active sites,but also establishes a favorable electronic and mass-transport environment.Furthermore,the roll-pressing-based integrated design streamlines electrode construction,reinforces interfacial bonding,and significantly enhances mechanical stability.Density functional theory(DFT)calculations show that oxygen functional groups initiate hydrogen bonding interaction and promote charge enrichment,which improves the activity of the cathode and facilitates intermediate adsorption/desorption in oxygen reduction and evolution reactions processes.As a result,the integrated air cathode-based rechargeable zinc-air batteries(RZABs)achieve a high specific capacity of 811 mAh g-1.It also performs well in quasi-solid-state RZABs and silicon-air batteries systems across a wide temperature range,demonstrating strong adaptability and application potential.This study provides a scalable and cost-effective design strategy for high-performance carbon-based air cathodes,offering new insights into advancing durable and practical metal-air energy systems.
基金Supported by National Natural Science Foundation of China(Grant Nos.52202431,52172353).
摘要Based on the demands for crashworthiness and lightweight in the passive safety of transportation vehicles,metal-fiber reinforced polymer(FRP)hybrid thin-walled tubes(MFHTWTs)integrate the toughness,strength and lightweight of two distinct material characteristics.MFHTWTs can achieve energy absorption through the coupling of material plastic deformation and fracture,demonstrating significant engineering value in passive safety.This review provides a comprehensive examination of the crashworthiness topology optimization of MFHTWTs,aiming to demonstrate that a deeply integrated approach combining topology and parameter opti-mization can realize an optimal design method for MFHTWTs,thereby maximizing the functional utilization of limited material.Firstly,the review highlights the crashworthiness topology optimization methods(CTOMs)based on thin-walled structures.With a particular focus on metal,the review discusses both the practical ap-plicability and limitations of CTOMs under crash conditions.Additionally,based on the methodology of the equivalent static load method(ESLM),the review emphasizes that topology optimization methods considering continuous fiber paths and multi-material interface connections are also applicable to the crashworthiness op-timization of MFHTWTs.Furthermore,to couple structural parameters and configuration characteristics,in-tegrated topology optimization methods,including parameter optimization,are proposed to provide a valuable reference for the global optimization of MFHTWTs.Thus,these methods can establish the mapping relationship between key parameters and the structural energy absorption capacity.
基金supported in part by the National Natural Science Foundation of China(Grant No.62201032)the Fundamental Research Funds for the Central Universities(Grant No.FRF-TP-22-045A1)the Young Elite Scientists Sponsorship Program by BAST(Grant No.BYESS2023306).
摘要Recently,the technique of Integrated Sensing and Communication(ISAC)has gained great attentions and is expected to enable more advanced applications in cellular Vehicle-to-Everything(V2X).We intend to use the periodical reference signals,such as the synchronization signals in 5G,to actively detect targets at different azimuths via beam scanning.This paper mainly aims to optimize the beamwidth of these reference signals in cellular-V2X to facilitate both sensing and communication functions.Firstly,to address the issue of insufficient beam duration for accurate estimation of Doppler and the corresponding velocity as well,we combine multi-ple beam scanning cycles to support improved sensing performance.In this context,an innovative beamwidth optimization algorithm is proposed.Specifically,we deduce the Fisher Information Matrix(FIM)of the suffi-cient statistic associated with target azimuths.Considering the randomness of the target position,we build the objective function based on the expected trace of the derived FIM.Additionally,we formulate the constraint of channel coherent time for efficient communications.It is worth noting that this is the first time to optimize the beamwidth of periodic reference signals for joint communication and sensing in cellular V2X.Experimental results show that the optimal beamwidth varies with the steering azimuth of antennas and reaches its maximum when the beam points at the direction perpendicular to the roadside.
基金supported by the National Natural Science Foundation of China(No.12272240)the Liaoning Revitalization Talents Program,China(No.XLYC2203197)。
摘要Traditional dynamic analysis of mechanical structures,often limited to individual beams or plates,fails to fully capture their dynamic behaviors.In systems where space and mass are constrained,such as the battery support structures in electric aircraft,conventional absorbers and isolators are insufficient for effective vibration control.This study simplifies the battery support structure of electric aircraft as an integrated composite beam consisting of three interconnected beams,and investigated its structural dynamics properties and nonlinear vibration control under thermal conditions caused by battery heat.The nonlinear vibration control is performed using the Nitinol steel wire ropes(Ni Ti-ST),with nonlinear damping properties.The natural frequencies of system are determined using the Rayleigh-Ritz technique.Theoretical results are validated through both Finite Element Method(FEM)and hammer tests.Moreover,the dynamic equations are derived using the Lagrange method and discretized via the Galerkin Truncation Method(GTM).The Harmonic Balance Method(HBM)is used to evaluate the vibration responses of the integrated model,with further verification through the Runge-Kutta Method(RKM).The experiments are conducted to corroborate the theoretical analysis.The results show that the system frequency changes in stages with the increase of the stiffness of the integrated composite beam connection.Especially in the case of varying environments,as the temperature increases,the frequency of system will first increase to a certain maximum value and then gradually decrease.Furthermore,the NiTi-ST effectively reduces vibration in the integrated composite beam,particularly under varying temperatures and external excitations.
基金supported by the National Natural Science Foundation of China(Grants No.42277041 and 42077163)the Postdoctoral Science Foundation of China(Grant No.2022M720316).
摘要Microbial Cr(Ⅵ)reduction represents a cost-effective remediation approach,yet its application in continuous-flow systems is often constrained by toxicity inhibition and nutrient instability.This study developed an integrated microbial—phosphorus mineral—corncob(MPC)composite,which encapsulates Cr(Ⅵ)-reducing bacteria with slow-release nutrients within a hydrogel matrix.The MPC system sustained complete Cr(Ⅵ)removal at influent concentrations up to 160 mg/L in synthetic wastewater and 120 mg/L in actual groundwater,conditions under which conventional systems typically fail.Reactive transport modeling revealed that diffusion limitations create protective intraparticle concentration gradients,while density functional theory(DFT)calculations confirmed strong Cr(Ⅲ)sequestration and sustained phosphorus bioavailability.The estimated treatment cost is competitive with existing technologies.These findings demonstrate that engineering a protected self-sufficient microbial microenvironment can overcome critical stability limitations in Cr(Ⅵ)bioremediation.
基金Opening Project of the National Key Laboratory of Advanced Polymer Materials(Sichuan University)(sklpme2024-1-08)。
摘要Silicon(Si)anodes offer an ultrahigh theoretical capacity but face two major barriers to commercialization:severe structural degradation caused by significant volume changes and sluggish ion transport kinetics due to the discontinuous ionic conductance of conventional binders.To address these challenges,we develop a fully integrated Si anode using sulfonated graphene(SG)as a dual-functional ion-conductive and mechanical reinforcing framework within a conventional carboxymethyl cellulose/styrene-butadiene rubber(CMC/SBR)binder.Thermally activated reactions during electrode fabrication establish covalent sulfonate ester bonds between SG and CMC,elastic carboxylate crosslinks between SBR and CMC,and chemical anchoring between the binders and Si particles,which are all further reinforced by hydrogen bonding.This multi-bonding network not only dissipates mechanical stress and maintains electronic connectivity via embedded C65 carbon but also significantly enhances Li+transport through high intrinsic ionic conductivity of SG,facilitating the formation of a stable solid electrolyte interphase(SEI).The resulting Si@CMC/SBR/SG anode delivers a high initial capacity of 3513.2 mAh g-1and retains 77%capacity after 500 cycles at 2 A g-1.It achieves 762 mAh g-1at 4 A g-1and practical areal capacities exceeding 4 mAh cm-2.Full-cell tests with NCM811 cathodes confirm 86.2%capacity retention after 100 cycles.This work demonstrates a pragmatic and scalable integration paradigm for durable,high-energy-density Si anodes.
基金supported in part by the National Natural Science Foundation of China(Grant No.62201034)the Beijing Municipal Natural Science Foundation(Grant No.L212004-03).
摘要For more accessible and advanced health monitoring,the Body Area Network(BAN)design with semantic technologies offers efficient information sensing and communication in smart healthcare Artificial Intelligence of Things(AIoT).To address the critical challenges of effective communication and reduction of data transmission pressure in AIoT-BAN,a hybrid BAN system is proposed which enhances information processing and communication capabilities by leveraging semantic understanding and multimodal processing.It incorporates a semantic communication and sensing fusion framework,offloading based on the human Body Coupled Communication(BCC)channel,and multimodal semantic information integration to reduce data transmission pressure.The proposed method offers effective inclusive smart healthcare and daily health maintenance for the general public.
摘要Transformer models face significant computational challenges in private inference(PI).Existing optimization methods often rely on isolated techniques,neglecting joint structural and operational improvements.We propose IG-3D,a unified framework that integrates structured compression and operator approximation through accurate importance assessment.Our approach first evaluates attention head importance using Integrated Gradients(IG),offering greater stability and theoretical soundness than gradient-based methods.We then apply a threedimensional optimization:(1)structurally pruning redundant attention heads;(2)replacing Softmax with adaptive polynomial approximation to avoid exponential computations;(3)implementing layer-wise GELU substitution to accommodate different layer characteristics.A joint thresholdmechanism coordinates compression across dimensions under accuracy constraints.Experimental results on the GLUE benchmark show that our method achieves an average 2.9×speedup in inference latency and a 50%reduction in communication cost,while controlling the accuracy loss within 2.3%,demonstrating significant synergistic effects and a superior accuracy-efficiency trade-off compared to single-technique optimization strategies.