检索结果-内蒙古大学图书馆

Deep-Unfolded Massive Grant-Free Transmission in Cell-Free Wireless Communication Systems

ieee TRANSACTIONS ON signal processing 2025年 73卷 1094-1109页

作者： Sun, Gangle Cao, Mengyao Wang, Wenjin Xu, Wei Studer, Christoph Southeast Univ Natl Mobile Commun Res Lab Nanjing 210096 Peoples R China Purple Mt Labs Nanjing 211100 Peoples R China Swiss Fed Inst Technol Dept Informat Technol & Elect Engn CH-8092 Zurich Switzerland

Grant-free transmission and cell-free communication are vital in improving coverage and quality-of-service for massive machine-type communication. This paper proposes a novel framework of joint active user detection, channel estimation, and data detection (JACD) for massive grant-free transmission in cell-free wireless communication systems. We formulate JACD as an optimization problem and solve it approximately using forward-backward splitting. To deal with the discrete symbol constraint, we relax the discrete constellation to its convex hull and propose two approaches that promote solutions from the constellation set. To reduce complexity, we replace costly computations with approximate shrinkage operations and approximate posterior mean estimator computations. To improve active user detection (AUD) performance, we introduce a soft-output AUD module that considers both the data estimates and channel conditions. To jointly optimize all algorithm hyper-parameters and to improve JACD performance, we further deploy deep unfolding together with a momentum strategy, resulting in two algorithms called DU-ABC and DU-POEM. Finally, we demonstrate the efficacy of the proposed JACD algorithms via extensive system simulations.

关键词： signal processing algorithms Channel estimation Wireless communication Approximation algorithms Vectors Optimization machine learning algorithms Training Sparse matrices Decoding Active user detection cell-free communication channel estimation data detection deep unfolding grant-free transmission massive machine-type communication

来源：评论

学校读者我要写书评

暂无评论

A novel amalgamation of pre-processing technique and CNN model for accurate classification of power quality disturbances

引用

ELECTRICAL ENGINEERING 2025年第4期107卷 5187-5206页

作者： Soni, Prity Mishra, Pankaj Mondal, Debasmita Birla Inst Technol Dept Elect & Elect Engn Ranchi 835215 India Thapar Inst Engn & Technol Dept Elect & Instrumentat Engn Patiala 147004 Punjab India

This work presents an innovative framework that combines the recurrence plots (RP) method with ResNet-50 (a convolutional neural network (CNN)) to autonomously extract relevant features for classifying multiple power quality disturbances for a power signal using the support vector machine. The ResNet-50 is employed to extract the most discriminated features from the two-dimensional images obtained from 1-D signals using the RP method. The work investigates synthetic power quality disturbances, including nine single disturbances, eight double disturbances, and seven triple disturbances. The validation of the proposed framework is conducted on the Standard ieee 5-bus system under diverse fault scenarios, demonstrating the method's efficiency and reliability. The statistical stability of the proposed framework is assessed using two non-parametric tests, the Sign test and the Wilcoxon test, to ensure the reliability of the results. Furthermore, the proposed work is compared with other advanced CNN models and pre-processing techniques, highlighting its superior performance and effectiveness.

关键词： Power quality disturbances signal processing Image processing Feature extraction Deep learning Classification

来源：评论

学校读者我要写书评

暂无评论

Formant Tracking by Combining Deep Neural Network and Linear Prediction

IEEE OPEN JOURNAL OF SIGNAL PROCESSING

引用

ieee OPEN JOURNAL OF signal processing 2025年 6卷 222-230页

作者： Kadiri, Sudarsana Reddy Huang, Kevin Hagedorn, Christina Byrd, Dani Alku, Paavo Narayanan, Shrikanth Univ Southern Calif Los Angeles CA 90007 USA CUNY Coll Staten Isl Grad Ctr New York NY 10017 USA Aalto Univ Espoo 02150 Finland

Formant tracking is an area of speech science that has recently undergone a technology shift from classical model-driven signal processing methods to modern data-driven deep learning methods. In this study, these two domains are combined in formant tracking by refining the formants estimated by a data-driven deep neural network (DNN) with formant estimates given by a model-driven linear prediction (LP) method. In the refinement process, the three lowest formants, initially estimated by the DNN-based method, are frame-wise replaced with local spectral peaks identified by the LP method. The LP-based refinement stage can be seamlessly integrated into the DNN without any training. As an LP method, the study advocates the use of quasiclosed phase forward-backward (QCP-FB) analysis. Three spectral representations are compared as DNN inputs: mel-frequency cepstral coefficients (MFCCs), the spectrogram, and the complex spectrogram. Formant tracking performance was evaluated by comparing the proposed refined DNN tracker with seven reference trackers, which included both signal processing and deep learning based methods. As evaluation data, ground truth formants of the Vocal Tract Resonance (VTR) corpus were used. The results demonstrate that the refined DNN trackers outperformed all conventional trackers. The best results were obtained by using the MFCC input for the DNN. The proposed MFCC refinement (MFCC-DNNQCP-FB) reduced estimation errors by 0.8 Hz, 12.9 Hz, and 11.7 Hz for the first (F1), second (F2), and third (F3) formants, respectively, compared to the Deep Formants refinement (DeepF(QCP-FB)). When compared to the model-driven KARMA tracking method, the proposed refinement reduced estimation errors by 2.3 Hz, 55.5 Hz, and 143.4 Hz for F1, F2, and F3, respectively. A detailed evaluation across various phonetic categories and gender groups showed that the proposed hybrid refinement approach improves formanttracking performance across most test conditions.

关键词： Estimation Spectrogram signal processing Artificial neural networks Training Computational modeling Accuracy Data models Predictive models Overfitting Formant tracking linear prediction MFCCs spectrogram machine learning deep learning

来源：评论

学校读者我要写书评

暂无评论

LSHIM: Low-Power and Small-Area Inexact Multiplier for High-Speed Error-Resilient Applications

引用

ieee JOURNAL ON EMERGING AND SELECTED TOPICS IN CIRCUITS AND SYSTEMS 2025年第1期15卷 94-104页

作者： Izadi, Azin Jamshidi, Vahid Shahid Bahonar Univ Kerman Dept Comp Engn Kerman *** Iran

Numerical computations in various applications can often tolerate a small degree of error. In fields such as data mining, encoding algorithms, image processing, machine learning, and signal processing where error resilience is crucial approximate computing can effectively replace precise computing to minimize circuit delay and power consumption. In these contexts, a certain level of error is permissible. Multiplication, a fundamental arithmetic operation in computer systems, often leads to increased circuit delay, power usage, and area occupation when performed accurately by multipliers, which are key components in these applications. Thus, developing an optimal multiplier represents a significant advantage for inexact computing systems. In this paper, we introduce a novel approximate multiplier based on the Mitchell algorithm. The proposed design has been implemented using the Cadence software environment with the TSMC 45nm standard-cell library and a supply voltage of 1.1V. Simulation results demonstrate an average reduction of 31.7% in area, 46.8% in power consumption, and 36.1% in circuit delay compared to previous works. The mean relative error distance (MRED) for the proposed method is recorded at 2.6%.

关键词： Approximation algorithms signal processing algorithms Circuits Accuracy Delays Performance evaluation Power demand Arithmetic Adders Energy consumption Approximate multiplier approximate computing Mitchell's multiplication algorithm energy-consumption area delay

来源：评论

学校读者我要写书评

暂无评论

Empowering Federated learning With Implicit Gossiping: Mitigating Connection Unreliability Amidst Unknown and Arbitrary Dynamics

引用

ieee TRANSACTIONS ON signal processing 2025年 73卷 766-780页

作者： Xiang, Ming Ioannidis, Stratis Yeh, Edmund Joe-Wong, Carlee Su, Lili Northeastern Univ Dept ECE Boston MA 02115 USA Carnegie Mellon Univ Dept ECE Pittsburgh PA 15213 USA

Federated learning is a popular distributed learning approach for training a machine learning model without disclosing raw data. It consists of a parameter server and a possibly large collection of clients (e.g., in cross-device federated learning) that may operate in congested and changing environments. In this paper, we study federated learning in the presence of stochastic and dynamic communication failures wherein the uplink between the parameter server and client i is on with unknown probability p(i)(t) in round t. Furthermore, we allow the dynamics of p(i)(t) to be arbitrary. We first demonstrate that when the p(i)(t)'s vary across clients, the most widely adopted federated learning algorithm, Federated Average (FedAvg), experiences significant bias. To address this observation, we propose Federated Postponed Broadcast (FedPBC), a simple variant of FedAvg. It differs from FedAvg in that the parameter server postpones broadcasting the global model to the clients with active uplinks till the end of each training round. Despite uplink failures, we show that FedPBC converges to a stationary point of the original non-convex objective. On the technical front, postponing the global model broadcasts enables implicit gossiping among the clients with active links in round t. In spite of the time-varying nature of pti, we can bound the perturbation of the global model dynamics using techniques to control gossip-type information mixing errors. Extensive experiments have been conducted on real-world datasets over diversified unreliable uplink patterns to corroborate our analysis.

关键词： Servers Uplink Federated learning Vehicle dynamics Training Stochastic processes Distance learning Computer aided instruction Smart phones Convergence communication failures gossiping non-convex optimization fault-tolerance

来源：评论

学校读者我要写书评

暂无评论

CSI Compression Method With Dual Differential Feedback for Next-Generation Wi-Fi Networks

引用

ieee WIRELESS COMMUNICATIONS LETTERS 2025年第2期14卷 475-478页

作者： Barannikov, Andrey Levitsky, Ilya Loginov, Vyacheslav Khorov, Evgeny Russian Acad Sci Inst Informat Transmiss Problems Wireless Networks Lab Moscow 141171 Russia

The multiple-input multiple-output (MIMO) technology improves Wi-Fi throughput by increasing the number of antennas. However, with more antennas and developing coordinated MIMO operations, the amount of channel state information (CSI) increases. As the CSI-induced overhead limits the performance of Wi-Fi networks, this letter proposes a new low-complexity dual differential CSI compression method that significantly reduces this overhead. Experimental and simulation results in diverse scenarios show that the proposed algorithm overcomes state-of-the-art algorithms of similar complexity, reducing the feedback size by up to 40%.

关键词： Wireless fidelity Array signal processing signal to noise ratio Matrix decomposition Time-domain analysis Standards Frequency-domain analysis Accuracy Throughput machine learning CSI compression MU-MIMO Wi-Fi differential feedback

来源：评论

学校读者我要写书评

暂无评论

Multi-Task Spiking Neural Network for Simultaneous Vapor Recognition and Concentration Estimation

引用

ieee ACCESS 2025年 13卷 29847-29863页

作者： Sartori Locatelli, Pedro Ait Fares, Salma Martini Colombo, Dalton El-Sankary, Kamal Freund, Michael S. Dalhousie Univ Dept Elect & Comp Engn Halifax NS B3J 1Z1 Canada InterTalk Crit Informat Syst Dartmouth NS B3B 0J5 Canada Univ Fed Minas Gerais Dept Elect Engn BR-31270901 Belo Horizonte Brazil Dalhousie Univ Dept Chem Halifax NS B3H 4R2 Canada

Neural networks have been instrumental in advancing machine olfaction systems, greatly enhancing their ability to process olfactory information. As the drive to integrate sensing and signal processing on chip continues, recent advancements emphasize the need for simple yet efficient pattern recognition systems. Spiking Neural Networks (SNNs) have gained significant attention in the realm of machine olfaction for their high computational efficiency and biological realism, closely mirroring the human olfactory system's approach to processing odors. Many studies have successfully applied SNN models for vapor classification;however, limited research exists on their use for concentration estimation. This article not only addresses this gap but also introduces a simple three-layer, multi-task shared SNN that simultaneously performs vapor recognition and concentration estimation using carbon black-polymer composite sensor arrays. Multi-task learning using SNNs proves particularly beneficial in this context by utilizing shared computations across tasks to optimize resource usage. Experimental results demonstrate that integrating both tasks into a single network enhances overall computational efficiency and reduces model complexity without sacrificing performance. Remarkably, this proof-of-concept design achieves a 33% reduction in the total number of neurons while performing 48% fewer synaptic operations per forward pass compared to two separate single-task SNNs. Logic synthesis further reveals that the shared approach consumes approximately 37% less power and occupies 37% less silicon area, showcasing the potential of multi-task learning with SNNs to propel artificial olfaction systems.

关键词： Polymers Multitasking Estimation Olfactory Sensor arrays Sensors Carbon Biology Biological neural networks Concentration estimation electronic nose gas recognition machine olfaction multi-task learning spiking neural network spiking neural network

来源：评论

学校读者我要写书评

暂无评论

Seismic Denoising Based on Dictionary learning With Double Regularization for Random and Erratic Noise Attenuation

引用

ieee TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 2025年 63卷

作者： Shekhar, Nakka Tejaswi, Dokku James, Abin Kuruguntla, Lakshmi Dodda, Vineela Chandra Mandpura, Anup Kumar Chinnadurai, Sunil Elumalai, Karthikeyan SRM Univ Dept Elect & Commun Engn Amaravati 522502 Andhra Pradesh India Koneru Lakshmaiah Educ Fdn KLEF Dept Elect & Commun Engn Vaddeswaram 522302 Andhra Pradesh India Amrita Vishwa Vidyapeetham Dept Elect & Commun Engn Amaravati 522503 Andhra Pradesh India Delhi Technol Univ Dept Elect Engn New Delhi 110042 India

In seismic data processing, denoising is one of the essential steps to identifying the earth's subsurface layer information. The noise present in the seismic data is categorized into two types: random and erratic noise. The random noise is distributed uniformly over the seismic data. The erratic noise attenuation is always challenging due to the unknown distribution of high-amplitude peaks over seismic data. The existing double sparsity dictionary learning (DSDL) method performs with analytical and adaptive transforms;both the transforms include iterative algorithms with K-singular-value decomposition (SVD);it is computationally costly, and the dictionary is initialized with trained data. To address these limitations, we propose a novel method of dictionary learning with double regularization (DLDR) to denoise both random and erratic noise from seismic data. In double regularization, we used with l(1)-norm and nuclear norm. The denoised data is applied to the alternating direction method of multipliers (ADMMs) to improve denoising while preserving the signal features from seismic data while reducing the computational cost. We evaluated the performance of the proposed method using signal-to-noise ratio (SNR), mean squared error (MSE), and local similarity map. The numerical results demonstrated that the proposed method resulted in higher SNR, lower MSE, and less signal leakage from seismic data. The method gives precise interpretation from the denoised seismic data.

关键词： Noise Noise reduction machine learning Dictionaries Attenuation Matching pursuit algorithms Iterative methods Filtering Continuous wavelet transforms signal to noise ratio & ell (1)-norm alternating direction method of multiplier (ADMM) dictionary learning double regularization erratic noise nuclear norm random noise

来源：评论

学校读者我要写书评

暂无评论

Graph Neural Networks With Adaptive Structures

引用

ieee JOURNAL OF SELECTED TOPICS IN signal processing 2025年第1期19卷 181-194页

作者： Zhang, Zepeng Lu, Songtao Huang, Zengfeng Zhao, Ziping ShanghaiTech Univ Sch Informat Sci & Technol Shanghai 201210 Peoples R China Ecole Polytech Fed Lausanne Lab Intelligent Maintenance & Operat Syst CH-1015 Lausanne Switzerland IBM Thomas J Watson Res Ctr Yorktown Hts NY 10598 USA Fudan Univ Sch Data Sci Shanghai 200433 Peoples R China ShanghaiTech Univ Sch Informat Sci & Technol Shanghai 201210 Peoples R China

Graph neural networks (GNNs) have made significant progress in various machine learning tasks. Despite their success, many existing GNN models are shown to be vulnerable to adversarial attacks, creating a stringent need to build robust GNN architectures. In this work, we introduce a novel interpretable message passing scheme over adaptive graph structures (ASMP) to enhance the resilience of GNNs against graph structural attacks. The ASMP layers are constructed through optimization steps that concurrently optimize node features and graph structures, allowing the message passing process to be conducted across dynamically adjusted graphs at different layers. This adaptability enables ASMP to handle noisy or perturbed graph structures more effectively, enhancing robustness. We also establish the theoretical convergence properties for the ASMP scheme. By integrating ASMP with neural networks, we introduce a new class of GNNs with adaptive structures (ASGNNs). Extensive experiments on semi-supervised node classification tasks demonstrate that ASGNN outperforms the state-of-the-art GNN architectures regarding classification accuracy when subjected to various adversarial attack scenarios.

关键词： Graph neural networks Message passing Adaptation models Perturbation methods Robustness Purification Optimization Noise measurement Heuristic algorithms Electronic mail Adversarial attacks graph neural network graph signal denoising graph signal processing graph structure learning

来源：评论

学校读者我要写书评

暂无评论

Audio-Based Kinship Verification Using Age Domain Conversion

引用

ieee signal processing LETTERS 2025年 32卷 301-305页

作者： Sun, Qiyang Akman, Alican Jing, Xin Milling, Manuel Schuller, Bjorn W. Imperial Coll London Dept Comp GLAM London SW7 2AZ England Tech Univ Munich CHI Chair Hlth Informat MRI D-81675 Munich Germany MDSI Munich Data Sci Inst D-85748 Munich Germany MCML Munich Ctr Machine Learning D-80539 Munich Germany

Audio-based kinship verification (AKV) is important in many domains, such as home security monitoring, forensic identification, and social network analysis. A key challenge in the task arises from differences in age across samples from different individuals, which can be interpreted as a domain bias in a cross-domain verification task. To address this issue, we design the notion of an "age-standardised domain" wherein we utilise the optimised CycleGAN-VC3 network to perform age-audio conversion to generate the in-domain audio. The generated audio dataset is employed to extract a range of features, which are then fed into a metric learning architecture to verify kinship. Experiments are conducted on the KAN_AV audio *** results demonstrate that the method markedly enhances the accuracy of kinship verification, while also offering novel insights for future kinship verification research.

关键词： Feature extraction Training Generators Videos Generative adversarial networks Vectors Hidden Markov models Data models Time-frequency analysis Recording Audio kinship verification deep learning GAN voice conversion

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：