检索结果-内蒙古大学图书馆

Do algorithms and Barriers for Sparse Principal Component Analysis Extend to Other Structured Settings?

IEEE TRANSACTIONS ON signal processing 2024年 72卷 3187-3200页

作者： Wang, Guanyi Lou, Mengqi Pananjady, Ashwin Natl Univ Singapore Dept Ind Syst Engn & Management Singapore 117576 Singapore Georgia Inst Technol H Milton Stewart Sch Ind & Syst Engn Atlanta GA 30332 USA Georgia Inst Technol Sch Elect & Comp Engn Atlanta GA 30332 USA

We study a principal component analysis problem under the spiked Wishart model in which the structure in the signal is captured by a class of union-of-subspace models. This general class includes vanilla sparse PCA as well as its variants with graph sparsity. With the goal of studying these problems under a unified statistical and computational lens, we establish fundamental limits that depend on the geometry of the problem instance, and show that a natural projected power method exhibits local convergence to the statistically near-optimal neighborhood of the solution. We complement these results with end-to-end analyses of two important special cases given by path and tree sparsity in a general basis, showing initialization methods and matching evidence of computational hardness. Overall, our results indicate that several of the phenomena observed for vanilla sparse PCA extend in a natural fashion to its structured counterparts.

关键词： Principal component analysis Computational modeling Estimation Indexes Vectors signal processing algorithms Iterative methods structured sparsity nonconvex iterative optimization computational hardness

来源：评论

学校读者我要写书评

暂无评论

Optimizing Convolutions for Deep Learning Inference on ARM Cortex-M Processors

引用

IEEE INTERNET OF THINGS JOURNAL 2024年第15期11卷 26203-26219页

作者： Macia-Lillo, Antonio Barrachina, Sergio Fabregat, German Dolz, Manuel F. Univ Alacant Dept Technol Informat & Computat San Vicente Del Raspeig 03690 Spain Univ Jaume 1 Dept Comp Sci & Engn Castellon de La Plana 12071 Spain

We perform a series of optimizations on the convolution operator within the ARM common microcontroller units software interface standard for neural network (CMSIS-NN) library to improve the performance of deep learning tasks on Arduino development boards equipped with ARM Cortex-M4 and M7 microcontrollers. To this end, we develop custom microkernels that efficiently handle the internal computations required by the convolution operator via the lowering approach and the direct method, and we design two techniques to avoid register spilling. We also take advantage of all the RAM on the Arduino boards by reusing it as a scratchpad for the convolution filters. The integration of these techniques into CMSIS-NN, when invoked by TensorFlow Lite for microcontrollers for quantised versions of VGG, SqueezeNet, ResNet, and MobileNet-like convolutional neural networks enhances the overall inference speed by a factor ranging from 1.13xto 1.50x .

关键词： Program processors Random access memory Registers Optimization Convolution Inference algorithms signal processing algorithms ARM Cortex-M common microcontroller units software interface standard for neural network (CMSIS-NN) convolution operator deep learning edge computing high performance microcontrollers

来源：评论

学校读者我要写书评

暂无评论

Joint Spectrum Sensing and DOA Estimation Based on a Resource-Efficient Sub-Nyquist Array Receiver

引用

IEEE TRANSACTIONS ON signal processing 2024年 72卷 5354-5370页

作者： Liu, Liang Zhang, Zhan Zhang, Xinyun Wei, Ping An, Jiancheng Li, Hongbin Univ Elect Sci & Technol China Sch Informat & Commun Engn Natl Key Lab Blind Signal Proc Chengdu 611731 Peoples R China Singapore Univ Technol & Design SUTD Engn & Prod Dev Pillar Singapore 487372 Singapore Stevens Inst Technol Dept Elect & Comp Engn Hoboken NJ 07030 USA

As the demand for wireless communication continues to surge, spectrum congestion becomes more severe. Compressive spectrum sensing with joint frequency and Direction Of Arrival (DOA) estimation is instrumental to enable efficient spectrum utilization in ultra-wideband scenarios. To reduce hardware complexity, this paper proposes a resource-efficient array undersampling structure which is distinctive in that each array element only connects to one branch of the Modulated Wideband Converter (MWC), and the modulated signals in different branches have identical periods but different waveforms. The proposed structure integrates spatial sampling by array elements and temporal undersampling by the MWC. A signal model is developed for the proposed sampling structure, which can deal with more general scenarios, involving multiple subband signals, and cross-band signals. Joint spectrum sensing algorithms are proposed based on compressed sensing and subspace decomposition. Additionally, multi-resolution grid optimization strategy is designed to eliminate grid effect with low computational complexity. We also analyze the impact of structural parameters on algorithm performance, which reveals that the number of array elements determines the maximum number of signals that can be estimated within a subband, while the equivalent total number of channels of the reception system determines the maximum number of signals that the system can estimate. Our analysis shows that the proposed sampling structure offers a greater flexibility in structural parameter selection and system design. Finally, simulations show that under the condition of the same or similar average sampling rate, the proposed structure and corresponding methods can achieve higher spectrum and DOA estimation accuracy.

关键词： Estimation Sensors Receivers Hardware Arrays Direction-of-arrival estimation Costs Channel estimation signal processing algorithms Frequency estimation Spectrum sensing DOA estimation sub-Nyquist sampling

来源：评论

学校读者我要写书评

暂无评论

Distributed Distortion-Aware Beamforming Designs for Cell-Free mMIMO Systems

引用

IEEE JOURNAL OF SELECTED TOPICS IN signal processing 2025年第2期19卷 381-397页

作者： Liu, Mengzhen Li, Ming Liu, Rang Liu, Qian Dalian Univ Technol Sch Informat & Commun Engn Dalian 116024 Peoples R China Univ Calif Irvine Ctr Pervas Commun & Comp Irvine CA 92697 USA Dalian Univ Technol Sch Comp Sci & Technol Dalian 116024 Peoples R China

Cell-free massive multi-input multi-output (CF-mMIMO) systems have emerged as a promising paradigm for next-generation wireless communications, offering enhanced spectral efficiency and coverage through distributed antenna arrays. However, the non-linearity of power amplifiers (PAs) in these arrays introduce spatial distortion, which may significantly degrade system performance. This paper presents the first investigation of distortion-aware beamforming in a distributed framework tailored for CF-mMIMO systems, enabling pre-compensation for beam dispersion caused by nonlinear PA distortion. Using a third-order memoryless polynomial distortion model, the impact of the nonlinear PA on the performance of CF-mMIMO systems is firstly analyzed by evaluating the signal-to-interference-noise-and-distortion ratio (SINDR) at user equipment (UE). Then, we develop two distributed distortion-aware beamforming designs based on ring topology and star topology, respectively. In particular, the ring-topology-based fully-distributed approach reduces interconnection costs and computational complexity, while the star-topology-based partially-distributed scheme leverages the superior computation capability of the central processor to achieve improved sum-rate performance. Extensive simulations demonstrate the effectiveness of the proposed distortion-aware beamforming designs in mitigating the effect of nonlinear PA distortion, while also reducing computational complexity and backhaul information exchange in CF-mMIMO systems.

关键词： Nonlinear distortion Array signal processing Matrix decomposition Vectors signal processing algorithms Interference Backhaul networks Analytical models Radio frequency Phase distortion Cell-free nonlinear distortion beamforming distributed signal processing

来源：评论

学校读者我要写书评

暂无评论

An Efficient and Accurate RCS Reconstruction Technique Using Adaptive TLS-ESPRIT Algorithm

引用

IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS 2024年第1期23卷 49-53页

作者： Zhu, Fang-Yin Chai, Shui-Rong Zou, Yu-Feng He, Zhen-Xiang Guo, Li-Xin Xidian Univ Sch Phys Xian 710071 Peoples R China

In this letter, an adaptive total least squares-estimating signal parameter via rotational invariance techniques (TLS-ESPRIT) algorithm based on the analytical perspective is proposed to efficiently construct the scattering center model (SCM) and reconstruct the radar cross section (RCS) of a perfect electrical conductor (PEC) target. Compared with existing research, the proposed method does not require knowing the number of scattering centers, which means that it can be applied to the construction of SCMs for noncooperative targets. Compared with inverse synthetic aperture radar (ISAR) and statistical methods, this work strikes a balance between accuracy and efficiency through optimization of the core technology in the RCS reconstruction process. The effectiveness of the algorithm is verified by comparing the results derived by our method with those from the measurement.

关键词： Scattering Adaptation models signal processing algorithms Radar cross-sections Radar scattering Image reconstruction Statistical analysis Adaptive TLS-ESPRIT algorithm Hankel matrix one-dimensional (1-D) scattering center radar cross section (RCS) reconstruction

来源：评论

学校读者我要写书评

暂无评论

Generalized Exponentiated Gradient algorithms and Their Application to On-Line Portfolio Selection

引用

IEEE ACCESS 2024年 12卷 197000-197020页

作者： Cichocki, Andrzej Cruces, Sergio Sarmiento, Auxiliadora Tanaka, Toshihisa POLISH ACAD SCI SYST RES INST PL-01447 Warsaw Poland Nicolaus Copernicus Univ UMK Dept Informat PL-87100 Torun Poland Tokyo Univ Agr & Technol Inst Global Innovat Res Koganei Tokyo 1840012 Japan Riken AIP Tokyo 1030027 Japan Univ Seville Dept Teoria Senal & Comunicac Seville 41092 Spain Tokyo Univ Agr & Technol Dept Elect Engn & Comp Sci Koganei Tokyo 1848588 Japan

Stochastic gradient descent (SGD) and exponentiated gradient (EG) update methods are widely used in signal processing and machine learning. This study introduces a novel family of generalized Exponentiated Gradient updates (EGAB) derived from the alpha-beta (AB) divergence regularization. The EGAB framework provides enhanced flexibility for processing data with varying distributions, thanks to the tunable hyperparameters of the AB divergence. We explore the applicability of these updates in online portfolio selection (OLPS) for financial markets with the goal of developing algorithms that achieve high risk-adjusted returns, even under relatively high transaction costs. The proposed EGAB algorithms are developed using constrained gradient optimization with regularization terms, demonstrating their versatility in OLPS by unifying the directional search of various algorithms and enabling interpolation between them. Our analysis and extensive computer simulations reveal that EGAB updates outperform existing OLPS algorithms, delivering good results on several performance metrics, such as cumulative return, average excess return, Sharpe ratio, and Calmar ratio, especially when transaction costs are significant. In conclusion, this study introduces a new family of exponentiated gradient updates and demonstrates their flexibility and effectiveness through extensive simulations across a wide range of real-world financial datasets.

关键词： Vectors signal processing algorithms Standards Machine learning algorithms Additives Portfolios Machine learning Cost function Probability distribution Costs Alpha-beta divergences exponentiated gradient algorithms on-line portfolio selection

来源：评论

学校读者我要写书评

暂无评论

Cooperative Positioning of Wireless Networks in Complex Propagation Environments

引用

IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS 2024年第10期42卷 2877-2889页

作者： Jiang, Peiyue Gu, Xiaobo Zhou, Haibo Guangdong Univ Technol Sch Integrated Circuits Guangzhou 510006 Peoples R China Nanjing Univ Sch Elect Sci & Engn Nanjing 210008 Peoples R China

Cooperative positioning in wireless networks has attracted great attention in recent years, as many applications require the exact location of all member nodes. The pairwise distance between the member nodes is conventionally constructed as an Euclidean Distance Matrix (EDM) for subsequent location estimation. In this paper, we address the problem of cooperative positioning in complex propagation environments, which results in an incomplete EDM. We proposed an improved EDM recovery algorithm based on low tank matrix completion (LRMC), which makes use of the sensor correlation by Laplacian and trace minimization. In addition, we derive a semi-definite relaxation estimator to localize the unknown sensors. Simulations are conducted to evaluate the performance of the proposed algorithm and the results show that the proposed method outperforms existing ones in both matrix completion and positioning accuracy.

关键词： Distance measurement Correlation Wireless sensor networks signal processing algorithms Position measurement Minimization Accuracy Cooperative positioning graph Laplacian low rank matrix completion manifold learning semi-definite programming

来源：评论

学校读者我要写书评

暂无评论

Robust CFAR Detector Based on CLEAN for Sidelobe Suppression

引用

IEEE SENSORS JOURNAL 2024年第9期24卷 14930-14942页

作者： Zhu, Jiachen Huang, Xiaohong Liu, Jiayu Deng, Zhenmiao North China Univ Sci & Technol Sch Artificial Intelligence Tangshan 063210 Peoples R China Hebei Key Lab Ind Intelligent Percept Tangshan 063210 Peoples R China Sun Yat Sen Univ Sch Elect & Commun Engn Shenzhen 518107 Peoples R China Sun Yat Sen Univ Sch Elect & Commun Engn Guangzhou 510275 Peoples R China

In response to challenges posed by abnormal interference, such as target sidelobes within the reference window, this article proposes a sidelobe suppression-constant false alarm rate (SS-CFAR) detector. Drawing on the CLEAN concept, the SS-CFAR is designed for target detection in indoor environments with multiple targets using millimeter (mm)-wave radar. It uses second-order difference statistics combined with the Monte Carlo (MC) algorithm and iterative censoring (IC) to separate target signals from background clutter in radar echoes. In addition, sidelobe suppression (SS) for the target is achieved by reconstructing the target's echo information and the CLEAN concept. The simulation and actual experimental results demonstrate that SS-CFAR delivers superior detection performance, obviating the need for prior knowledge of anomalies or manual setting of parameters such as the number of reference and guard cells, which traditionally depend on experiential adjustments. Compared with mean class and ordered statistics (OS) class CFAR detectors, SS-CFAR transcends reference window constraints. It accurately estimates background clutter, enhancing detection capabilities in multitarget environments. This innovative approach not only showcases superior detection performance and adaptability but also paves the way for the broader application of mm-wave radar in complex indoor scenarios.

关键词： Clutter signal processing algorithms Object detection Radar detection Detectors Millimeter wave radar Sensors CLEAN constant false alarm rate (CFAR) interfering targets millimeter (mm)-wave radar reference window sidelobe suppression (SS) target detection

来源：评论

学校读者我要写书评

暂无评论

Robust Iterative Sparse ML-Based Approaches for CES Data Model: Application to DOA Estimation

引用

IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS 2024年第6期60卷 8635-8650页

作者： Abeida, Habti Glentis, George-Othon Alsharef, Mohammad Li, Jian Taif Univ Coll Engn Dept Elect Engn Taif 21944 Saudi Arabia Univ Peloponnese Dept Informat & Telecommun Tripolis 22100 Greece Univ Florida Dept Elect & Comp Engn Gainesville FL 32611 USA

SCM is widely used in the design of array signal processing algorithms that perform well for CCG distributed samples. However, SCM-based DOA estimation algorithms are inaccurate for CES data models. This article aims to extend the conventional SCM-based iterative sparse SMLA for DOA estimation to encompass more general circular CES distributed observations. These extended RSMLA use robust CES-based ML-estimators of the CM instead of SCM, or more generally M-estimators. Joint and interleaved computation of M-estimators and the RSMLA are also proposed [denoted as IRSMLA algorithms] when the size of the observed CES vector, M, is of the same order as the samples size, N, or when N < M. To improve DOA estimation accuracy and eliminate grid mismatch, refined approaches (referred to as RSMLA-RML) that combines the power-based RSMLA and robust ML (RML) methods are also proposed, in which an alternative RML cost function is minimized over a scalar DOA parameter. Performance comparisons are made between the developed approaches and existing DOA estimation algorithms as well as CRB in terms of sample size, SNR, DOA separation, signal source correlation parameters, and computational cost, demonstrating superior performance over existing schemes for heavy-tailed CES distributed observations. Moreover, it is experimentally demonstrated that the proposed algorithms perform effectively even when the encountered source signals are coherent, despite being derived under the assumption of uncorrelated sources.

关键词： Direction-of-arrival estimation signal processing algorithms Accuracy Maximum likelihood estimation signal to noise ratio Multiple signal classification Vectors Circular complex elliptically symmetric (CES) distributions direction of arrival (DOA) estimation M-estimators stochastic maximum likelihood approaches (SMLA) algorithms sparse representation

来源：评论

学校读者我要写书评

暂无评论

Fast Equiconcept Search on Graph Formal Context by Attribute Topology

引用

IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING 2024年第2期11卷 1469-1479页

作者： Zhang, Tao Li, Shaoze Xue, Zaifa Lu, Xin Yanshan Univ Sch Informat Sci & Engn Qinhuangdao 066000 Peoples R China Key Lab Informat Transmiss & Signal Proc Hebei Pro Qinhuangdao 066000 Peoples R China

The equiconcept, as an emerging topic in formal concept analysis, provides new ideas for analyzing the behavior in social computing. However, it takes too much time for calculation because of the high complexity. In this article, a fast equiconcept search method on graph formal context by attribute topology is proposed to improve computational efficiency. First, attribute topology is optimally represented as a graph context attribute topology (GcAT) model on the graph formal context. Then, it is proved that there is a one-to-one match between the maximal complete polygon in GcAT and the equiconcept in the graph formal context. Finally, combined with the existing attribute topology formal concept searching algorithm, the fast equiconcept search is completed. The experimental results show that by the proposed method not only all equiconcepts are searched, but also the time consumption is significantly reduced compared with the other methods over seven datasets. In addition, with the size of the dataset increasing, the computational efficiency of GcAT model improves more obviously, which indicates its superiority. All of that demonstrate the proposed method has greatly improved the efficiency of equiconcept search through GcAT on graph formal context.

关键词： Topology Lattices Social networking (online) Computational modeling Formal concept analysis Visualization signal processing algorithms Equiconcept formal concept analysis fast equiconcept search graph formal context graph context attribute topology

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：