检索结果-内蒙古大学图书馆

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Xu, Zhe Lei, Zhipei Gou, Dingyong Wu, Yanlin Zhang, Liwen Li, Cong ZTE Corporation Shenzhen518057 China State Key Laboratory of Mobile Network and Mobile Multimedia Technology Shenzhen518055 China

ISBN: (纸本)9798350368741

Image deblurring is dedicated to restoring blurry images resulting from camera shake or target motion into high-quality sharp images. Recent work has made notable progress in image deblurring, but few studies have focused on the role of high-frequency information in this task. Hence, an efficient Edge-aware Laplacian Pyramid network (ELPNet) is proposed for image deblurring. Specifically, we introduce a reversible Laplacian pyramid decomposition and reconstruction mechanism within the deblurring network, guiding the reconstruction of high-frequency information. Additionally, we present a novel Large-Kernel convolution Hybrid Attention Block (LKHAB) that leverages re-parameterization to effectively integrate channel-wise and spatial-invariant features with a larger receptive field. We also introduce an Edge-Aware Merge Block (EAMB) that combines Laplacian and Scharr operators with difference convolutions to create various learnable edge gradient convolutions. The EAMB can elegantly capture the edge information of features, and through re-parameterization, enables to merging of different edge gradient convolutions into a vanilla convolution during inference. Experimental results show that the proposed method achieves state-of-the-art performance with significantly reduced parameters and computation. © 2025 IEEE.

关键词： attention edge-aware Image deblurring Laplacian pyramid re-parameterization

来源：评论

学校读者我要写书评

暂无评论

Energy Consumption Minimization for NOMA-Based Secure UAV-MEC network

引用

China Communications 2025年第3期22卷 202-216页

作者： Zhang Hao Huang Yuzhen Zhang Zhi Lu Xingbo School of Information and Communication Engineering Beijing University of Posts and TelecommunicationsBeijing 100876China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing 100876China The Academy of Military Sciences of PLA Beijing 100091China

Applying non-orthogonal multiple access(NOMA)to the mobile edge computing(MEC)network supported by unmanned aerial vehicles(UAVs)can improve spectral efficiency and achieve massive user access on the basis of solving computing resource constraints and coverage ***,the UAV-enabled network has a serious risk of information leakage on account of the openness of wireless *** paper considers a UAV-MEC secure network based on NOMA technology,which aims to minimize the UAV energy *** achieve the purpose while meeting the security and users’latency requirements,we formulate an optimization problem that jointly optimizes the UAV trajectory and the allocation of network *** that the original problem is non-convex and multivariate coupled,we proposed an effective algorithm to decouple the nonconvex problem into independent user relation coefficients and subproblems based on successive convex approximation(SCA)and block coordinate descent(BCD).The simulation results showcase the performance of our optimization scheme across various parameter settings and confirm its superiority over other benchmarks with respect to energy consumption.

关键词： MEC NOMA resource optimization secure transmission trajectory optimization UAV

来源：评论

学校读者我要写书评

暂无评论

Electromagneticwave property inspired radio environment knowledge construction and artificial intelligence based verification for6G digital twin channel

引用

Frontiers of Information technology & Electronic Engineering 2025年第2期26卷 260-277页

作者： Jialin WANG Jianhua ZHANG Yutong SUN Yuxiang ZHANG Tao JIANG Liang XIA State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing 100876China China Mobile Research Institute Beijing 100053China

As the underlying foundation of a digital twin network(DTN),digital twin channel(DTC)can accurately depict the electromagnetic wave propagation in the air interface to support the DTN-based 6G wireless *** electromagnetic wave propagation is affected by the environment,constructing the relationship between the environment and radio wave propagation is the key to implementing *** the existing methods,the environmental information inputted into the neural network has many dimensions,and the correlation between the environment and the channel is unclear,resulting in a highly complex relationship construction *** solve this issue,we propose a unified construction method of radio environment knowledge(REK)inspired by the electromagnetic wave property to quantify the propagation contribution based on easily obtainable location *** effective scatterer determination scheme based on random geometry is proposed which reduces redundancy by 90%,87%,and 81%in scenarios with complete openness,impending blockage,and complete blockage,*** also conduct a path loss prediction task based on a lightweight convolutional neural network(CNN)employing a simple two-layer convolutional structure to validate REK’s *** results show that only 4 ms of testing time is needed with a prediction error of 0.3,effectively reducing the network complexity.

关键词： Digital twin channel Radio environment knowledge(REK)pool Wireless channel Environmental information Interpretable REK construction Artificial intelligence based knowledge verification

来源：评论

学校读者我要写书评

暂无评论

Coded Beam Training for RIS-Assisted Wireless Communications

引用

IEEE Transactions on Wireless Communications 2025年第5期24卷 3854-3867页

作者： Chen, Yuhao Dai, Linglong Tsinghua University State Key Laboratory of Space Network and Communications Department of Electronic Engineering Beijing100084 China Massachusetts Institute of Technology Department of Electrical Engineering and Computer Science CambridgeMA02139 United States

Reconfigurable intelligent surface (RIS) is considered as one of the key technologies for future 6G communications. To fully unleash the performance of RIS, accurate channel state information (CSI) is crucial. Beam training is widely utilized to acquire the CSI. However, before aligning the beam correctly to establish stable connections, the signal-to-noise ratio (SNR) at UE is inevitably low, which reduces the beam training accuracy. To deal with this problem, we exploit the coded beam training framework for RIS systems, which leverages the error correction capability of channel coding to improve the beam training accuracy under low SNR. Specifically, we first extend the coded beam training framework to RIS systems by decoupling the base station-RIS channel and the RIS-user channel. For this framework, codewords that accurately steer to multiple angles is essential for fully unleashing the error correction capability. In order to realize effective codeword design in RIS systems, we then propose a new codeword design criterion, based on which we propose a relaxed Gerchberg-Saxton (GS) based codeword design scheme by considering the constant modulus constraints of RIS elements. In addition, considering the two dimensional structure of RIS, we further propose a dimension reduced encoder design scheme, which can not only guarentee a better beam shape, but also enable a stronger error correction capability. Simulation results reveal that the proposed scheme can realize effective and accurate beam training in low SNR scenarios. © 2002-2012 IEEE.

关键词： Error correction

来源：评论

学校读者我要写书评

暂无评论

FSMM: An Efficient Matrix Multiplication Accelerator Supporting Flexible Sparsity 24

FSMM: An Efficient Matrix Multiplication Accelerator Support...

引用

43rd International Conference on Computer-Aided Design, ICCAD 2024

作者： Qiao, Yuxuan Yang, Fan Zhang, Yecheng Xiong, Xiankui Yao, Xiao Yao, Haidong State Key Laboratory of Integrated Chips and Systems Fudan University Shanghai China School of Microelectronics Fudan University Shanghai China State Key Laboratory of Mobile Network and Mobile Multimedia Technology ZTE Corporation Shenzhen China

ISBN: (纸本)9798400710773

Sparse matrix multiplication is a critical operation in deep learning. However, matrix sparsity leads to irregular data flow, which would degrade the efficiency of matrix multiplication. Traditional accelerators, equipped with additional hardware units to address this issue, often experience the issue of low hardware utilization. Furthermore, N : M structured sparsity and corresponding hardware architectures face challenges such as accuracy degradation, limited flexibility, and restricted applicability. In this paper, we propose a Flexible Sparse Matrix Multiplication Accelerator (FSMM), which can improve the efficiency of sparse matrix multiplication through both algorithmic-level and hardware-level optimizations. At the algorithmic level, we propose the matrix-matrix multiplication with block-level outer production and fine-grained matrix reordering algorithm. The algorithm balances the sparsity of each column of a matrix block, which improves matrix compression, balances the load, and speeds up computation. This algorithm reduces storage by 8.2% ~ 85.9%. At the hardware-level, we introduce a flexible architecture for matrix multiplication. It selects the most suitable data path to complete matrix multiplication based on the sparsity of the reordered matrix. FSMM achieves a speedup of 1.90× ~ 16.18× over Systolic Array and 1.70× ~ 2.87× over the existing TSTC approach. © 2024 Copyright is held by the owner/author(s).

关键词： Fast response computer systems

来源：评论

学校读者我要写书评

暂无评论

MMH-FE:AMulti-Precision and Multi-Sourced Heterogeneous Privacy-Preserving Neural network Training Based on Functional Encryption

引用

Computers, Materials & Continua 2025年第3期82卷 5387-5405页

作者： Hao Li Kuan Shao Xin Wang Mufeng Wang Zhenyong Zhang The State Key Laboratory of Public Big Data College of Computer Science and TechnologyGuizhou UniversityGuiyang550025China Key Laboratory of Computing Power Network and Information Security Ministry of EducationShandong Computer Science CenterQilu University of Technology(Shandong Academy of Sciences)Jinan250014China China Industrial Control Systems Cyber Emergency Response Team Beijing100040China

Due to the development of cloud computing and machine learning,users can upload their data to the cloud for machine learning model ***,dishonest clouds may infer user data,resulting in user data *** schemes have achieved secure outsourced computing,but they suffer from low computational accuracy,difficult-to-handle heterogeneous distribution of data from multiple sources,and high computational cost,which result in extremely poor user experience and expensive cloud computing *** address the above problems,we propose amulti-precision,multi-sourced,andmulti-key outsourcing neural network training ***,we design a multi-precision functional encryption computation based on Euclidean ***,we design the outsourcing model training algorithm based on a multi-precision functional encryption with multi-sourced ***,we conduct experiments on three *** results indicate that our framework achieves an accuracy improvement of 6%to 30%.Additionally,it offers a memory space optimization of 1.0×2^(24) times compared to the previous best approach.

关键词： Functional encryption multi-sourced heterogeneous data privacy preservation neural networks

来源：评论

学校读者我要写书评

暂无评论

An Efficient Coherent Integration Method for Highly Maneuvering Target Detection by Exploiting Range-Azimuth Coupling Relationships Between Subsegments

引用

IEEE Sensors Journal 2025年第9期25卷 15456-15476页

作者： Wan, Jun Jiang, Xinyu Liu, Chengyuan Li, Dong Xie, Xiaoran Sang, Qinglin Chen, Zhanye Chongqing University School of Microelectronics and Communication Engineering Chongqing Key Laboratory of Space Information Network and Intelligent Information Fusion Chongqing400044 China China Academy of Electronics and Information Technology Beijing100041 China State Key Laboratory of Millimeter Waves Institute of Electromagnetic Space Southeast University Nanjing210096 China

Arbitrary-order range migrations (RMs) and Doppler frequency migrations (DFMs) caused by complicated motions of highly maneuvering targets (HMTs) pose challenges for coherent integration (CI) in radar remote sensing and detection applications. Complicated Doppler frequency ambiguity improves the challenges of CI for HMTs. The optimal generalized Radon-Fourier transform (GRFT) suffers from a large computational cost, and nonlinear transform-based methods are affected by severe performance loss. To tackle these difficulties, an efficient segmented CI method for HMT detection via the range-azimuth coupling relationship between subsegments is proposed. First, inspired by the advantage of subaperture processing, a segmentation operation is introduced to remove intrasegment RMs and DFMs. Second, a high-order compensation function is presented to remove the high-order intersegment RMs and DFMs. Third, by exploiting the range-azimuth coupling relationship between subsegments, a residual migration compensation function is used to correct the residual intersegment RM. Thus, a well result of HMT's CI is achieved. The proposed method handles arbitrary-order RMs and DFMs, complex Doppler ambiguity, and ambiguous speed sidelobes (BSSLs) in low signal-to-noise ratio (SNR) circumstances and multitarget cases due to its linear transform property. In comparison to the optimal GRFT, the proposed approach substantially lowers computational cost while showing similar performance. Then, the presented approach can be implemented by parallel processing. It is advantageous to enhance real-time processing efficiency. Several real radar data are analyzed to confirm the effectiveness of the developed method. © 2001-2012 IEEE.

关键词： Signal to noise ratio

来源：评论

学校读者我要写书评

暂无评论

Effects of visual rotation axis and clockwise/counterclockwise directions on circular vection in virtual reality 24

Effects of visual rotation axis and clockwise/counterclockwi...

引用

2024 International Conference on Virtual Reality, Image and Signal Processing, ICVISP 2024

作者： Xu, Xiaotian Song, Zhe Zhang, Kexin Xu, Yan Zhang, Yuankun Fan, Xiaoya Pu, Fang Key Laboratory of Human Motion Analysis and Rehabilitation Technology of the Ministry of Civil Affairs Beihang University Beijing China Key Laboratory for Ubiquitous Network and Service Software of Liaoning Province Dalian University of Technology Liaoning Dalian China State Key Laboratory of Virtual Reality Technology and System Beihang University Beijing China

ISBN: (纸本)9798400710926

Vection, an illusion of self-motion induced by visual stimulus frequently experienced in virtual reality, significantly impacts user experiences in gaming, driving simulation, and rehabilitation training. User often experience circular vection, an illusion of self-rotation. However, the effects of visual rotation axes and rotation directions on circular vection remain elusive. We employ a 3 (rotation axis: sagittal, coronal, vertical) × 2 (rotation direction: clockwise, counterclockwise) repeated-measures design. Twenty participants were recruited. They were exposed to a virtual visual rotating drum at 60 °/s to induce circular vection. The real-time estimation of perceived self-motion rotation velocity (PSRV) was obtained. Results indicated that during the 30-second constant rotation, the group-level PSRV showed an initial rapid increase followed by a slower rise. A significant interaction effect between rotation axis and direction was revealed. The rotation direction showed no significant effect on PSRV under sagittal and vertical conditions. However, under coronal condition, the PSRV was significantly higher when rotating counterclockwise, compared to rotation clockwise. In addition, when rotating counterclockwise, the highest PSRV appeared under sagittal condition, with the lowest under the coronal condition and the middle under the vertical condition. These findings provide valuable insights for designing VR scenarios with better user experience by vection manipulation. © 2024 Copyright held by the owner/author(s).

关键词： Virtual reality

来源：评论

学校读者我要写书评

暂无评论

Towards Space-Based Computing Infrastructure network: Development Trends, network Architecture, Challenges Analysis, and key Technologies

arXiv

引用

arXiv 2025年

作者： Kuang, Linling Sun, Jiachen Zhang, Jin Cui, Huanxi Liu, Kai Department of Electronic Engineering Tsinghua University Beijing100084 China State Key Laboratory of Space Network and Communications Tsinghua University Beijing100084 China Beijing National Research Center for Information Science and Technology Tsinghua University Beijing100084 China

As one of the most promising hotspots in the 6G era, space remote sensing information networks play a key and irreplaceable role in areas such as emergency response and scientific research, and are expected to foster remote sensing data processing into the next generation of killer applications. However, due to the inability to deploy ground communication stations at scale and the limited data transmission window, the traditional model for transmitting space data back to ground stations faces significant challenges in terms of timeliness. To address this problem, we focus on the emerging paradigm of on-orbit space data processing, taking the first step toward building a space-based computing infrastructure network. Specifically, we propose a hierarchical space-based computing network architecture that integrates the space-based cloud constellation system, the remote sensing constellation system, the network operation control center, the orchestration data center and the user access portal, offering a detailed description of their functionalities. Next, we analyze three scientific challenges: the characterization and virtualization of multidimensional heterogeneous resources, the efficient orchestration of multidimensional heterogeneous resources for tasks with varying priorities, and the rapid sharing of multidimensional heterogeneous resources to address burst tasks or system failures. Finally, we discuss key technologies to address the aforementioned challenges and highlight promising research priorities for the future. Copyright © 2025, The Authors. All rights reserved.

关键词： Data centers

来源：评论

学校读者我要写书评

暂无评论

Dynamic Staleness Control for Asynchronous Federated Learning in Decentralized Topology 18th

Dynamic Staleness Control for Asynchronous Federated Learni...

引用

18th International Conference on Wireless Artificial Intelligent Computing Systems and Applications, WASA 2024

作者： Ma, Qianpiao Liu, Jianchun Jia, Qingmin Zhou, Xiaomao Hu, Yujiao Xie, Renchao Purple Mountain Laboratories Jiangsu Nanjing China School of Computer Science and Technology University of Science and Technology of China Anhui Hefei China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China

ISBN: (纸本)9783031714665

Decentralized federated learning (DFL) has emerged as a promising paradigm for distributed machine learning over edge nodes (i.e., workers) without relying on a centralized parameter server. Most existing DFL researches rely on synchronous communication among workers. However, due to edge heterogeneity and dynamic network conditions, synchronous DFL mechanisms may suffer from inefficient model training and poor scalability. Meanwhile, the existing asynchronous DFL (ADFL) mechanisms present the challenge of stale models among workers, leading to diminished training quality, especially on Non-IID data. In this paper, we propose a novel staleness-aware ADFL (SA-ADFL) mechanism, aiming to realize a trade-off between model training efficiency and quality by dynamic staleness control. Specifically, we provide rigorous theoretical proof for SA-ADFL and formulate the worker scheduling problem to minimize total model training time under flexible long-term staleness constraints. Then we decompose the original round-coupled problem into a series of single-round sub-problems by leveraging the Lyapunov optimization, enabling efficient worker selection to minimize training time in each round while ensuring staleness queue stability. Experimental results demonstrate that our SA-ADFL accelerates model training by approximately 52.9% while maintaining equivalent model training accuracy compared with the state-of-the-art mechanisms. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：