检索结果-内蒙古大学图书馆

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Feng Xue Thierry Blu Jiaqi Liu Xia Ai National Key Laboratory of Science and Technology on Test Physics and Numerical Mathematics Beijing 100076 China Dept. of Electronic Engineering The Chinese University of Hong Kong Shatin N.T. Hong Kong

ISBN: (纸本)9781538646595

Recently, total variation (TV)-based regularization has become a standard technique for signal denoising. The reconstruction quality is generally sensitive to the value of regularization parameter. In this work, based on Cham-bolle's algorithm, we develop two data-driven optimization schemes based on minimization of Stein's unbiased risk estimate (SURE)-statistically equivalent to mean squared error (MSE). First, we propose a recursive evaluation of SURE to monitor the estimation error during Chambolle's iteration;the optimal value is then identified by the minimum SURE. Second, for fast optimization, we perform alternating update between regularization parameter and solution within Chambolle's iteration. We exemplify the proposed methods with both 1-D and 2-D signal denoising. numerical experiments show that the proposed methods lead to highly accurate estimate of regularization parameter and nearly optimal denoising performance.

关键词： Signal denoising total variation (TV) Stein's unbiased risk estimate (SURE) Chambolle's algorithm

来源：评论

学校读者我要写书评

暂无评论

The Effects of Base Shape on Cone Scattering Feature

The Effects of Base Shape on Cone Scattering Feature

引用

International Applied Computational Electromagnetics Society Symposium (ACES)

作者： Yongtao Shui Shengjun Zhang Hang Song Lei Mu Xianglin Zhang Yuning Zhang Yichun Cui Weidong Wang Mingliang Wang Ligen Cai Yinfang Xu Song Chai Jiaqi Liu Chao Dong Fengju Wang Zhengyu Yong Chunshou Shao Minghao Wang Zhaohui Qi Yuan Sun Chenxin Sun Beijing Institute of Space Long March Vehicle Beijing China National Key Laboratory of Science & Technology on Test Physics and Numerical Mathematics Beijing China Institute of Survival Technology effectiveness Evaluation of Flying Vehicle Beijing China

Recent days, radar cross section reduction(RCSR) of objects have attracted much interest as more and more wideband radar were used to detect low observable weapon platforms. Among the low observation technologies, shaping plays an important role. This paper focused on the effect of basement shape on the scattering of a cone-like object, and some basement shape better for low radar cross section design for around tip incident angle were analyzed.

关键词： Geometry Radar cross-sections Shape Surface waves Weapons Electromagnetic scattering Nose

来源：评论

学校读者我要写书评

暂无评论

Modulation recognition of emitter signals based on SDAE-ODCNN

Modulation recognition of emitter signals based on SDAE-ODCN...

引用

IET International Radar Conference (IET IRC 2020)

作者： X. Wang J. Gao L. Gao F. Wang Harbin Engineering University College of Information and Communication Engineering Harbin China Beijing Institute of Space Long March Vehicle National Key Laboratory of Science and Technology on Test Physics and Numerical Mathematics Beijing China

In order to improve the recognition rate of emitter signals modulation under low signal noise ratio (SNR), an end-to-end method based on the stacked denoising autoencoder (SDAE) - one dimensional convolutional neural network (ODCNN) is proposed in this paper. The SADE-ODCNN is designed in this method and trained through a two-step training. In the SDAE-ODCNN model, the encoder of the SDAE can extract the denoised and compressed features of the received emitter signals. Then the ODCNN can further extract features to realize the recognition. The simulation results show that the average recognition rate of the method can reach 0.94 under 0dB. Thus the proposed method has a good application in the field of emitter signals modulation recognition.

关键词： radar computing modulation feature extraction neural nets image denoising learning (artificial intelligence) image classification

来源：评论

学校读者我要写书评

暂无评论

A Compound THz Meta-surface Absorber with Polarization Insensitivity and Angular Stability

A Compound THz Meta-surface Absorber with Polarization Insen...

引用

IEEE Asia-Pacific Conference on Antennas and Propagation (APCAP)

作者： Kai Wang Hailiang Zhu Ganyu Liu Jinchao Mou Pei Zheng Gao Wei Northwestern Polytechnical University School of Electronics and Information Xi'an Shaanxi CN Beijing Research Institute of Telemetry Beijing CN The National Key Laboratory of Science and Technology on Test Physics and Numerical Mathematics Beijing CN

In this paper, a wideband terahertz meta-surface absorber with ultra-thin thickness is proposed. The meta-surface unit cell is composed of four-square unit cells in two different sizes. The the absorption higher than 90% from 3.24 to 3.98THz is achieved which can be regarded as a perfect terahertz absorber. Due to the symmetrical layout of meta-surface unit cell, the absorption spectrum can be maintained for any linear polarized terahertz wave, and the absorptance can be kept higher than 80% while the incident angle increases before 45°. Furthermore, the thermal response is calculated to be 0.23 $\mathbf{K} \cdot{\mathbf{m}\mathbf{m}^{2}/ \upmu \mathbf{W}}$ that well suited the requirements of FPA detectors.

关键词： Sensitivity Absorption Layout Detectors Stability analysis Compounds Thermal stability

来源：评论

学校读者我要写书评

暂无评论

Digital Nonlinear Self-Interference Cancellation Using Sw-Gru-Tpa

SSRN

引用

SSRN 2024年

作者： Zhao, Zhongkai Xia, Shijie Gao, Lu Zheng, Pei Jiang, Yilin College of Information and Communication Engineering Harbin Engineering University Harbin150000 China National Key Laboratory of Science and Technology on Test Physics and Numerical Mathematic Beijing100876 China

In order to enhance spectrum utilization and decrease transmission delay, the 6G system adopts full-duplex(FD) mode. Simultaneous operation of the same frequency inevitably causes the receiver to receive the signal sent by the transmitter, leading to serious nonlinear self-interference due to the nonlinear distortion generated by the power amplifier(PA). This paper proposes hybrid layer structures SW-GRU-TPA for the implementation of digital self-interference cancellation(SIC), combined with Sliding Window (SW), Gated Recurrent Unit (GRU) , and Temporal Pattern Attention (TPA).The model can not only effectively utilize the influence of historical information on the signal, but also capture the spatial and temporal dependencies in the data to fit the nonlinear characteristics of the PA and the delay effect in multipath transmission. Meanwhile, by utilizing the L1 norm model pruning, the model parameters can be optimized, leading to improved processing efficiency, enhanced practicality, and increased convenience in real-world applications. Through the verification of the actual data, the performance of the SW-GRU-TPA network has improved by 7.95 dB, 7.79 dB, 3.45 dB and 3.13 dB respectively, compared to the RVTDNN, SW-GRU, HCRDNN and SW-GRU-AM *** the TPA is compressed by 42.17%, the ICR is only reduced by 1.58%. © 2024, The Authors. All rights reserved.

关键词： Power amplifiers

来源：评论

学校读者我要写书评

暂无评论

Doffnet: A Distant Objects Rgb-T Semantic Segmentation Method Based on Cross-Level Interaction Fusion Network

SSRN

引用

SSRN 2024年

作者： Chen, Yu Li, Xiang Luan, Chao Hou, Weimin Liu, Haochen Xue, Lian Zhu, Zihui Zhang, Jianqi Liu, Delian Wu, Xin Wei, Linfang Jian, Chaochao Li, Jinze Academy of Military Sciences School of Physics and Optoelectronic Engineering Xidian University Xi’an710071 China Beijing Institute of Control & Electronics Technology Beijing100038 China National Key Laboratory of Science and Technology on Test Physics & Numerical Mathematics Beijing100076 China

RGB-T segmentation represents an innovative approach driven by advancements in multi-spectral detection, poised to replace traditional RGB segmentation methods. An effective cross-modality feature fusion module is essential for this technology;simultaneously, the precise segmentation of distant objects is likewise a significant challenge. Building on these two focuses, we propose an end-to-end distant object feature fusion network (DOFFNet) for RGB-T segmentation. Initially, we introduce a cross-level interaction fusion strategy (CLIF) and an inter-correlation fusion method (IFFM) in the encoder to enhance multi-scale feature expression and improve fusion accuracy. Subsequently, we propose a residual-dense pixel convolution (R-DPC) in the decoder with a trainable upsampling unit, which dynamically reconstructs information lost during encoding, particularly for distant objects whose features may vanish after pooling. Experimental results demonstrate that DOFFNet improves accuracy across five classes and effectively segments distant objects occupying 0.2-2% of total pixels, enhancing overall segmentation quality and detailed understanding of RGB-T scenes. © 2024, The Authors. All rights reserved.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Meta or select?* : —history, argument and prospective about MTMs and FSS

Meta or select?* : —history, argument and prospective about...

引用

International Applied Computational Electromagnetics Society Symposium (ACES)

作者： Shengjun Zhang Mingliang Wang Xia Ai Yichun Cui Xin Liu Jiaqi Liu Gang Meng Zhaohui Qi Yuan Sun Yunhui Tang Hui Xue National Key Laboratory of Science & Technology on Test Physics and Numerical Mathematics Beijing China The Institute of Effectiveness Evaluation of Flying Vehicle Beijing China School of materials science and technology Beijing University of Technology Beijing China Beijing Institute of system engineering Beijing China

In this paper, the development history of metamaterials (MTMs) and Frequency selective surfaces (FSS) was sketched, and the two conception was discussed. Furthermore, state of the art of both terms was introduced and ... 详细信息

关键词： Frequency selective surfaces Metamaterials Lattices Crystals physics Electromagnetics Sun

来源：评论

学校读者我要写书评

暂无评论

A Radar Signal Modulation Recognition Algorithm Based on Multi-branch Fusion Difference

A Radar Signal Modulation Recognition Algorithm Based on Mul...

引用

IEEE International Conference on Electronic Information and Communication technology (ICEICT)

作者： Sisi Jiang Jingpeng Gao Zhiye Jiang Fang Ye College of Information and Communication Engineering Harbin Engineering University Haibin China National Key Laboratory of Science and Technology on Test Physics and Numerical Mathematics Beijing Institute of Space Long March Vehicle Beijing China National Key Laboratory of Underwater Acoustic Technology College of Information and Communication Engineering Harbin Engineering University Harbin China

The time-frequency images of radar modulation signals are similar at low signal-to-noise ratios, which will lead to lower recognition accuracy. This paper proposes an algorithm based on multi-branch fusion differences to identify radar modulation signals. The algorithm designs a multi-branch fusion differential module based on the attention mechanism and differential thinking, to extract time-frequency image fine-grained features and improve the Convolution Neural Network characterization ability. After that, a dual-branch differential network is proposed by combining vanilla convolution and multi-branch fusion difference module. The network can fuse global and detailed features of time-frequency images, which can achieve radar signal modulation recognition. The simulation results show that the recognition accuracy of the algorithm can reach 93.17% at 0dB. Therefore, the algorithm can obtain high recognition accuracy and has high recognition performance.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Fast Design Method for Ultra Broadband Absorbers Using RC Equivalent Circuit Model

A Fast Design Method for Ultra Broadband Absorbers Using RC ...

引用

International Applied Computational Electromagnetics Society Symposium (ACES)

作者： Xianglin Kong Lei Zhao Fengjun Wang Shengjun Zhang Yixiang Fu Xiaoyu Pang Xiaochun Liu Xiaopeng Shen School of Information and Control Engineering China University of Mining and Technology Xu Zhou CA China National Key Laboratory of Science and Technology on Test Physics & Numerical Mathematical Beijing China Key Laboratory of Aeronautical Science and Technology for High Performance Electromagnetic Window Jinan China School of Materials and Physics China University of Mining and Technology Xuzhou China

An accuracy and efficient approach is proposed to fast design ultra-wideband microwave absorbers. The RC equivalent circuit model of absorbers is established according to impedance matching and transmission line theory. The model is suitable for low and high-frequency absorbers composed of multilayer square high impendence surface (HIS) and air layers of impendence matching. Absorbers with different parameters can be fast designed by utilizing the proposed model and genetic algorithm. The reflection coefficient of RC circuit and full-simulation are in good agreement, both of them are lower than -10 dB within 2-18 GHz or 4-50 GHz (fractional bandwidth of 160% or 170.37%), with the total thickness close to the Rozanov limit. An absorber is fabricated and measured to further verify the accuracy of the model. The measured results have a good agreement with the RC circuit model, which indicates the proposed method is an efficient way for fast designing ultra-broadband thin absorbers using double-layer (or multilayer) square HIS.

关键词： Microwave measurement Atmospheric modeling Transmission line measurements Nonhomogeneous media Microwave theory and techniques Microwave circuits Frequency measurement

来源：评论

学校读者我要写书评

暂无评论

Spectral–spatial attention fusion transformer for hyperspectral image classification

Spectral–spatial attention fusion transformer for hyperspec...

引用

IET International Radar Conference (IRC 2023)

作者： Xiangyu Ji Jingpeng Gao Geng Chen Lu Gao College of Information and Communication Engineering Harbin Engineering University Harbin People's Republic of China National Key Laboratory of Science and Technology on Test Physics and Numerical Mathematics Beijing Institute of Space Long March Vehicle Beijing People's Republic of China

ISBN: (数字)9781837240982

In view of the limited generalization and poor local feature information modeling ability of vision transformer model(ViT) in hyperspectral image classification, a Spectral-Spatial Attention Fusion Transformer (SAFT) method for hyperspectral image classification is proposed in this paper. The proposed method includes two key modules: Spectral-Spatial Feature Extraction Module (SSFE) and Multi-Head Attention Fusion Transformer Module (MAFT). The SSFE module is a lightweight network. It contains a 3D convolution block and a 2D convolution block, which can efficiently obtain local feature information. Based on the proposed attention fusion structure, the MAFT module designs an improved transformer encoder. It improves the generalization ability of the model on the premise of obtaining global information. In this paper, two datasets of HyRANK are used for experiments. The experimental results show that this method is superior to other HSI classification methods.

关键词：

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：