检索结果-南通市图书馆

On the size generalizibility of graph neural networks for learning resource allocation

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第4期67卷 239-254页

作者： Jiajun WU Chengjian SUN Chenyang YANG Beihang University school of electronics and Information Engineering

Size generalization is important for learning resource allocation policies in wireless systems with time-varying scales. If a neural network for learning a wireless policy is not generalizable to the size of its input, it has to be re-trained whenever the system scale changes, which hinders its practical use due to the unaffordable training costs. Graph neural networks(GNNs) have been shown with size generalization ability empirically when optimizing resource allocation. Yet, are GNNs naturally size generalizable? In this paper,we argue that GNNs are not always size generalizable for resource allocation. We find that the aggregation and activation functions of the GNNs for learning a class of wireless policies play a key role in their size generalization ability. We take the GNN with the mean aggregator, called mean-GNN, as an example to reveal a size generalization condition. To demonstrate how to satisfy the condition, we learn power and bandwidth allocation policies for ultra-reliable low-latency communications and show that selecting or pretraining the activation function in the output layer of mean-GNN can make the GNN size *** results validate our analysis and evaluate the performance of the learned policies.

关键词： size generalization graph neural networks resource allocation permutation equivariance

A comprehensive analysis of DAC-SDC FPGA low power object detection challenge

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第8期67卷 300-320页

作者： Jingwei ZHANG Guoqing LI Meng ZHANG Xinye CAO Yu ZHANG Xiang LI Ziyang CHEN Jun YANG school of electronics Science and Engineering Southeast University

The lower power object detection challenge(LPODC) at the IEEE/ACM Design Automation Conference is a premier contest in low-power object detection and algorithm(software)-hardware co-design for edge artificial intelligence, which has been a success in the past five years. LPODC focused on designing and implementing novel algorithms on the edge platform for object detection in images taken from unmanned aerial vehicles(UAVs), which attracted hundreds of teams from dozens of countries to participate. Our team SEUer has been participating in this competition for three consecutive years from 2020 to 2022 and obtained sixth place respectively in 2020 and 2021. Recently, we achieved the championship in 2022. In this paper, we presented the LPODC for UAV object detection from 2018 to 2022, including the dataset, hardware platform,and evaluation method. In addition, we also introduced and discussed the details of methods proposed by each year's top three teams from 2018 to 2022 in terms of network, accuracy, quantization method, hardware performance, and total score. Additionally, we conducted an in-depth analysis of the selected entries and results, along with summarizing representative methodologies. This analysis serves as a valuable practical resource for researchers and engineers in deploying the UAV application on edge platforms and enhancing its feasibility and reliability. According to the analysis and discussion, it becomes evident that the adoption of a hardware-algorithm co-design approach is paramount in the context of tiny machine learning(TinyML).This approach surpasses the mere optimization of software and hardware as separate entities, proving to be essential for achieving optimal performance and efficiency in TinyML applications.

关键词： tiny machine learning object detection convolutional neural networks algorithm-hardware co-design low power field programmable gate array

Magnetic field imaging with radio-frequency optically pumped magnetometers [Invited]

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Optics Letters 2024年第6期22卷 38-45页

作者： Xiyu Liu Junlong Han Wei Xiao Teng Wu Xiang Peng Hong Guo school of electronics Peking UniversityBeijing 100871China

High-sensitivity radio-frequency optically pumped magnetometers (RF-OPMs), working without cryogenic condition, play a critical role in magnetic field imaging(MFI) at low frequencies(e.g., 100 Hz to 1 MHz). We introduce the principle of operation and recent developments of RF-OPMs and focus on reviewing the MFI applications in magnetic induction tomography, ultralow-field magnetic resonance imaging, and magnetic particle imaging. For the applications of RF-OPMs, ranging from industrial monitoring to medical imaging and security screening, the unshielded and portable RF-OPMs(and RF-OPM array)techniques are still under the further development for detecting and scanning over the target object for accomplishing the final three-dimensional imaging, and thus extremely require the abilities of active compensation of the ambient magnetic field and sensor miniaturization in the future.

关键词： magnetic field imaging radio frequency optically pumped magnetometer

Three-dimensional topological crystalline insulator without spin-orbit coupling in nonsymmorphic photonic metacrystal

在线全文

学校读者我要写书评

暂无评论

Chinese Physics B 2024年第10期33卷 282-286页

作者： Zhide Yu Lingbo Xia school of Physics and electronics Hunan UniversityChangsha 410082China

By including certain point group symmetry in the classification of band topology,Fu proposed a class of threedimensionaltopological crystalline insulators(TCIs)without spin-orbit coupling in *** Fu’s model,surface states(ifpresent)doubly degenerate atГandM when time-reversal and C4 symmetries are *** analogs of Fu’s modelwith surface states quadratically degenerate atM are widely studied,while surface states with quadratic degeneracy atГare rarely *** this study,we propose a three-dimensional TCI without spin-orbit coupling in a judiciously designednonsymmorphic photonic *** surface states of photonic TCIs exhibit quadratic band degeneracy in the(001)surface Brillouin zone(BZ)center(Гpoint).The gapless surface states and their quadratic dispersion are protected by C4and time-reversal symmetries,which correspond to the nontrivial band topology characterized by Z_(2)topological ***,the surface states along lines fromГto the(001)surface BZ boundary exhibit zigzag feature,which is interpretedfrom symmetry perspective by building composite operators constructed by the product of glide symmetries with *** metacrystal array surrounded with air possesses high order hinge states with electric fields highlylocalized at the hinge that may apply to optical *** gapless surface states and hinge states reside in a cleanfrequency *** topological surface states emerge at the boundary of the metacrystal and perfect electric conductor(PEC),which provide a pathway for topologically manipulating light propagation in photonic devices.

关键词： topological crystalline insulator nonsymmorphic photonic metacrystal

Fractional-order heterogeneous memristive Rulkov neuronal network and its medical image watermarking application

在线全文

学校读者我要写书评

暂无评论

Chinese Physics B 2024年第5期33卷 306-314页

作者：丁大为牛炎张红伟杨宗立王金王威王谋媛 school of electronics and Information Engineering Anhui UniversityHefei 230601China

This article proposes a novel fractional heterogeneous neural network by coupling a Rulkov neuron with a Hopfield neural network(FRHNN),utilizing memristors for emulating neural *** study firstly demonstrates the coexistence of multiple firing patterns through phase diagrams,Lyapunov exponents(LEs),and bifurcation ***,the parameter related firing behaviors are described through two-parameter bifurcation ***,local attraction basins reveal multi-stability phenomena related to initial ***,the proposed model is implemented on a microcomputer-based ARM platform,and the experimental results correspond to the numerical ***,the article explores the application of digital watermarking for medical images,illustrating its features of excellent imperceptibility,extensive key space,and robustness against attacks including noise and cropping.

关键词： fractional order memristors Rulkov neuron medical image watermarking

Loyal wingman task execution for future aerial combat:A hierarchical prior-based reinforcement learning approach

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2024年第5期37卷 462-481页

作者： Jiandong ZHANG Dinghan WANG Qiming YANG Zhuoyong SHI Longmeng JI Guoqing SHI Yong WU school of electronics and Information Northwestern Polytechnical UniversityXi’an 710129China

In modern Beyond-Visual-Range(BVR)aerial combat,unmanned loyal wingmen are pivotal,yet their autonomous capabilities are *** study introduces an advanced control algorithm based on hierarchical reinforcement learning to enhance these capabilities for critical missions like target search,positioning,and relay *** on a dual-layer model,the algorithm’s lower layer manages basic aircraft maneuvers for optimal flight,while the upper layer processes battlefield dynamics,issuing precise navigational *** approach enables accurate navigation and effective reconnaissance for lead ***,our Hierarchical Prior-augmented Proximal Policy Optimization(HPE-PPO)algorithm employs a prior-based training,prior-free execution method,accelerating target positioning training and ensuring robust target *** paper also improves missile relay guidance and promotes the effective *** integrating this system with a human-piloted lead aircraft,this paper proposes a potent solution for cooperative aerial *** experiments demonstrate enhanced survivability and efficiency of loyal wingmen,marking a significant contribution to Unmanned Aerial Vehicles(UAV)formation control *** advancement is poised to drive substantial interest and progress in the related technological fields.

关键词： Beyond-visual-range Loyal wingmen Hierarchical prior-augmented proximal policy optimization Unmanned aerial vehicles Warfare

Air target intention recognition and causal effect analysis combining uncertainty information reasoning and potential outcome framework

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2024年第1期37卷 287-299页

作者： Yu ZHANG Fanghui HUANG Xinyang DENG Mingda LI Wen JIANG school of electronics and Information Northwestern Polytechnical UniversityXi’an 710129China

Recognizing target intent is crucial for making decisions on the ***,the imperfect and ambiguous character of battlefield situations challenges the validity and causation analysis of classical intent recognition *** with the challenge,a target intention causal analysis paradigm is proposed by combining with an Intervention Retrieval(IR)model and a Hybrid Intention Recognition(HIR)*** target data acquired by the sensors are modelled as Basic Probability Assignments(BPAs)based on evidence theory to create uncertain ***,the HIR model is utilized to recognize intent for a tested sample from uncertain ***,the intervention operator under the evidence structure is utilized to perform attribute intervention on the tested *** retrieval is performed in the sample database based on the IR model to generate the intention distribution of the pseudo-intervention samples to analyze the causal effects of individual sample *** simulation results demonstrate that our framework successfully identifies the target intention under the evidence structure and goes further to analyze the causal impact of sample attributes on the target intention.

关键词： Causal effect analysis Hybrid intention recognition Intervention retrieval Target intention Uncertainty reasoning

A novel Bayesian-based INS/GNSS integrated positioning method with both adaptability and robustness in urban environments

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2024年第6期37卷 205-218页

作者： Zhe YANG Hongbo ZHAO school of electronics and Information Engineering Beihang UniversityBeijing 100191China

Achieving higher accuracy positioning results in urban environments at a lower cost has been an important pursuit in areas such as autonomous driving and intelligent *** Inertial Navigation System and Global Navigation Satellite System(INS/GNSS)integrated navigation systems have been an important means of fulfilling the above quest due to the complementary error characteristics between INS and *** complex urban driving environment requires the system sufficiently adaptive to keep up with the time-varying measurement noise and sufficiently robust to cope with measurement outliers and prior ***,many efforts lack a balance between adaptability and *** this paper,a novel positioning method with both adaptability and robustness is proposed by coupling the Mahalanobis distance method,the Variational Bayesian method and the student’s t-distribution in one process(M-VBt method).This method is robust against non-Gaussian noise and priori uncertainties,plus adaptive against measurement noise uncertainty and time-varying *** field test results show that the M-VBt method(especially the Mahalanobis distance part)has significantly improved the system performance in the complex urban driving environment.

关键词： Urban environments Mahalanobis distance Adaptability Robustness Integrated navigation

Theoretical characterization of the adsorption configuration of pyrrole on Si(100)surface by x-ray spectroscopy

在线全文

学校读者我要写书评

暂无评论

Chinese Physics B 2024年第2期33卷 430-435页

作者：李好情明静姜志昂李海波马勇宋秀能 school of Physics and electronics Shandong Normal UniversityJinan 250358China

The possible configurations of pyrrole absorbed on a Si(100)surface have been investigated by x-ray photoelectron spectroscopy(XPS)and near-edge x-ray absorption fine structure(NEXAFS)*** C-1s XPS and NEXAFS spectra of these adsorption configurations have been calculated by using the density functional theory(DFT)method and fullcore hole(FCH)approximation to investigate the relationship between the adsorption configurations and the *** result shows that the XPS and NEXAFS spectra are structurally dependent on the configurations of pyrrole absorbed on the Si(100)*** with the XPS,the NEXAFS spectra are relatively sensitive to the adsorption configurations and can accurately identify *** NEXAFS decomposition spectra produced by non-equivalent carbon atoms have also been calculated and show that the spectral features vary with the diverse types of carbon atoms and their structural environments.

关键词： pyrrole silicon surface x-ray photoelectron spectroscopy(XPS) near-edge x-ray absorption fine structure(NEXAFS)