检索结果-内蒙古大学图书馆

Dual Class Token Vision Transformer for Direction of Arrival Estimation in Low SNR

IEEE SIGNAL PROCESSING LETTERS 2024年 31卷 76-80页

作者： Guo, Yu Zhang, Zhi Huang, Yuzhen Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China Acad Mil Sci PLA Beijing 100091 Peoples R China

In this letter, we propose a deep learning-based method for the direction of arrival (DOA) estimation in the low signal-to-noise ratio (SNR) scenario. Specifically, the DOA estimation is modeled as a multi-label classification task, and a novel dual class token Vision Transformer (DCT-ViT) is designed to fit it. Different from the classical ViT architecture with a single class token, the DCT-ViT includes two class tokens which are located at the beginning and end of the latent vector sequence, respectively. This architecture enables enhanced information mining and feature extraction from the array signal data in order to improve the accuracy of DOA estimation. Furthermore, a single DCT-ViT model can accommodate different source numbers by leveraging a training dataset with different numbers of sources. Simulation results illustrate that our proposed method outperforms existing methods in the low SNR scenario, including classical model-based and other deep learning-based methods.

关键词： Direction of arrival (DOA) estimation Vision Transformer (ViT) low signal-to-noise ratio (SNR) deep learning

来源：评论

学校读者我要写书评

暂无评论

A Concurrent Multibeam Passive Radar for Respiratory Detection From Multiple Subjects

引用

IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES 2025年

作者： Huang, Shuaiming Zhao, Heng Hong, Hong Guo, Fengxian Peng, Mugen Zhu, Xiaohua Nanjing Univ Sci & Technol Sch Elect & Opt Engn Nanjing 210094 Peoples R China Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing Peoples R China

The proliferation of the Internet of Things (IoT) technology has increased wireless interconnectivity, enhancing smart living environments. The millimeter-wave radar is prevalent for indoor positioning, and it consumes spectrum resources and risk interference in multiradar scenarios. In contrast, passive sensing leverages ambient electromagnetic signals, detecting targets through channel variations without occupying the spectrum. However, existing systems are constrained by hardware designs, thus resulting in limited environmental adaptability and poor sensing capabilities. In this article, a concurrent multibeam passive radar is proposed for localization and respiration detection of multiple human targets. Utilizing 16 independent digital receivers with a variable frequency local oscillator (LO), the radar adaptively senses ambient radio frequency signals and selects suitable ones for target detection. Upon identifying the positions of radiation sources and targets, the system generates multiple independent beams to capture motion, mitigating multipath echo interference from wide-beam antennas and enhancing environmental adaptability. Performance evaluation of the passive radar involved various experiments with simulated radiation sources, sliding platforms, and real human targets. The proposed passive radar system accurately localizes and distinguishes external illuminator from human targets, achieving an angular estimation error of sub-2(degrees) in 3 m. It also detects respiration signals through orthogonal baseband processing, with a maximum error of 0.2 beats per minute (BPM) and a median error below 0.03 BPM, ensuring high-precision respiratory monitoring.

关键词： Passive radar Sensors Radar Baseband Radar antennas Monitoring Interference Radar detection Wireless sensor networks Wireless fidelity Digital receiver multibeam passive radar noncontact vital sign detection

来源：评论

学校读者我要写书评

暂无评论

An Approximate Wave-Number Domain Expression for Near-Field XL-Array Channel

引用

IEEE TRANSACTIONS ON VEHICULAR technology 2025年第5期74卷 8267-8272页

作者： Xing, Hongbo Zhang, Yuxiang Zhang, Jianhua Xu, Huixin Liu, Guangyi Wang, Qixing Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China China Mobile Res Inst Future Res Lab Beijing 100053 Peoples R China

As Extremely large-scale array (XL-array) technology advances and carrier frequency rises, the near-field effects in communication are intensifying. In near-field conditions, channels exhibit a diffusion phenomenon in the angular domain, existing research indicates that this phenomenon can be leveraged for efficient parameter estimation and beam training. However, the channel model in angular domain lacks closed-form analysis, making the time complexity of the corresponding algorithm high. To address this issue, this paper analyzes the near-field diffusion effect in the wave-number domain, where the wave-number domain can be viewed as the continuous form of the angular domain. A closed-form approximate wave-number domain expression is proposed, based on the Principle of Stationary Phase. Subsequently, we derive a simplified expression for the case where the user's distance is much larger than the array aperture, which is more concise. Subsequently, we verify the accuracy of the proposed approximate expression through simulations and demonstrate its effectiveness using a beam training example. Results indicate that the beam training scheme, improved by the wave-number domain approximation model, can effectively estimate near-field user parameters and perform beam training using far-field DFT codebooks. Moreover, its performance surpasses that of existing DFT codebook-based beam training methods.

关键词： Training Discrete Fourier transforms Antennas Channel models Vectors Closed-form solutions Channel estimation Apertures Solid modeling Phased arrays Near-field extremely large-scale array (XL-array) wave-number domain the principle of stationary phase

来源：评论

学校读者我要写书评

暂无评论

Graph-based Anomaly Detection and Root Cause Analysis for Microservices in Cloud-Native Platform 14

Graph-based Anomaly Detection and Root Cause Analysis for Mi...

引用

14th IEEE International Conference on Sustainable Computing and Communications, SustainCom 2024

作者： Wang, Xinwei Liu, Xu Xu, Peng Du, Haoran State Key Lab of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China China Academy of Industrial Internet Beijing China

ISBN: (纸本)9798331521059

Due to the excellent characteristics of microservices architecture, such as high scalability, high fault tolerance, and ease of maintenance, and the flexible, efficient, and scalable deployment environment provided by cloud-native platforms, an increasing number of applications are being constructed based on microservices architecture. In such scenarios, the complexity of anomaly detection rises sharply due to the involvement of numerous interrelated microservices components and the complex invocation relationships among these components in each microservices application. Existing anomaly detection methods struggle to maintain sufficient accuracy in such complex environments. To address this challenge, this paper proposes the Graph-based Anomaly Detection and Root Cause Analysis (GADRCA) method for Microservices in Cloud-Native Platform. This method leverages graph structures to comprehensively considers the fundamental distribution characteristics of multiple resource consumption metrics involved in the runtime of microservices applications, such as response time, CPU usage, and memory usage. It also preserves the structural information of trace data within microservices to achieve more comprehensive anomaly detection. Upon identifying anomalous traces, the method determines the direction of anomaly propagation using the traces root cause analysis approach based on multiple metric anomaly types. Finally, a topological graph approach is employed to accurately pinpoint the root cause service. The proposed method's effectiveness in enhancing anomaly detection and root cause analysis has been demonstrated through a series of experiments. © 2024 IEEE.

关键词： Cloud platforms

来源：评论

学校读者我要写书评

暂无评论

CM-TTS: Enhancing Real Time Text-to-Speech Synthesis Efficiency through Weighted Samplers and Consistency Models

CM-TTS: Enhancing Real Time Text-to-Speech Synthesis Efficie...

引用

2024 Findings of the Association for Computational Linguistics: NAACL 2024

作者： Li, Xiang Bu, Fan Mehrish, Ambuj Li, Yingting Han, Jiale Cheng, Bo Poria, Soujanya State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications China Singapore University of Technology and Design Singapore

ISBN: (纸本)9798891761193

Neural Text-to-Speech (TTS) systems find broad applications in voice assistants, e-learning, and audiobook creation. The pursuit of modern models, like Diffusion Models (DMs), holds promise for achieving high-fidelity, real-time speech synthesis. Yet, the efficiency of multi-step sampling in Diffusion Models presents challenges. Efforts have been made to integrate GANs with DMs, speeding up inference by approximating denoising distributions, but this introduces issues with model convergence due to adversarial training. To overcome this, we introduce CM-TTS, a novel architecture grounded in consistency models (CMs). Drawing inspiration from continuous-time diffusion models, CM-TTS achieves top-quality speech synthesis in fewer steps without adversarial training or pre-trained model dependencies. We further design weighted samplers to incorporate different sampling positions into model training with dynamic probabilities, ensuring unbiased learning throughout the entire training process. We present a real-time mel-spectrogram generation consistency model, validated through comprehensive evaluations. Experimental results underscore CM-TTS's superiority over existing single-step speech synthesis systems, representing a significant advancement in the field. © 2024 Association for Computational Linguistics.

关键词： Efficiency

来源：评论

学校读者我要写书评

暂无评论

Multi-Domain Image-to-Image Translation with Cross-Granularity Contrastive Learning

引用

ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS 2024年第7期20卷 1-21页

作者： Fu, Huiyuan Liu, Jin Yu, Ting Wang, Xin Ma, Huadong Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing Key Lab Intelligent Telecommun Software & Beijing Beijing Peoples R China SUNY Stony Brook Dept Elect & Comp Engn Stony Brook NY USA

The objective of multi-domain image-to-image translation is to learn the mapping from a source domain to a target domain in multiple image domains while preserving the content representation of the source domain. Despite the importance and recent efforts, most previous studies disregard the large style discrepancy between images and instances in various domains, or fail to capture instance details and boundaries properly, resulting in poor translation results for rich scenes. To address these problems, we present an effective architecture for multi-domain image-to-image translation that only requires one generator. Specifically, we provide detailed procedures for capturing the features of instances throughout the learning process, as well as learning the relationship between the style of the global image and that of a local instance in the image by enforcing the cross-granularity consistency. In order to capture local details within the content space, we employ a dual contrastive learning strategy that operates at both the instance and patch levels. Extensive studies on different multi-domain image-to-image translation datasets reveal that our proposed method outperforms state-of-the-art approaches.

关键词： Image-to-image translation GAN cross-granularity contrastive learning multi-domain

来源：评论

学校读者我要写书评

暂无评论

Optimizing Space-Borne Computation: A Reliability Enhancement Framework for LEO Constellation 29

Optimizing Space-Borne Computation: A Reliability Enhancemen...

引用

29th IEEE International Conference on Parallel and Distributed Systems, ICPADS 2023

作者： Yang, Lipei Zhou, Ao Ma, Xiao Zhang, Yiran Wang, Shangguang Beijing University of Posts and Telecommunications The State Key Laboratory of Networking and Switching Technology China

ISBN: (纸本)9798350330717

Edge computing is extending the frontier of computation beyond terrestrial boundaries, with Low Earth Orbit (LEO) constellations emerging as the cutting-edge paradigm, aspiring to deliver ubiquitous computing capabilities globally. However, the unresolved service reliability issues within LEO constellations continue to barricade the realization of the vision for space computing. Terrestrial-native classical methods struggle to meet the stringent environmental conditions of space computing and often fail to satisfy the rigorous requirements of LEO constellations for failure response times and computational resource overhead. To address these challenges, we introduce the LEO Service Reliability Enhancement Framework (LSREF). LSREF employs the Variable Speed Replication technology to balance failure response times and computational resource overhead and adopts a decentralized design, more congruent with the characteristics of LEO constellations. Specifically, to counter the immense scale and high dynamism of LEO constellations, LSREF proposes orbital plane-based autonomous domains and leverages the Satellite Autonomous Backup Deployment Algorithm in conjunction with the Registry Polling Mechanism to enable autonomous decision-making for service backup strategies on each satellite. Our simulation experiments demonstrate that, compared to classical methods, LSREF reduces average failure response times by 10.76% and diminishes computational resource consumption by nearly 15.53%. © 2023 IEEE.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Efficient, Scalable, and Sustainable DNN Training on SoC-Clustered Edge Servers

引用

IEEE TRANSACTIONS ON MOBILE COMPUTING 2024年第12期23卷 14344-14360页

作者： Xu, Mengwei Xu, Daliang Lou, Chiheng Zhang, Li Huang, Gang Jin, Xin Liu, Xuanzhe Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China Peking Univ Sch Comp Sci Key Lab High Confidence Software Technol Minist Educ Beijing 100871 Peoples R China

In the realm of industrial edge computing, a novel server architecture known as SoC-Cluster, characterized by its aggregation of numerous mobile systems-on-chips (SoCs), has emerged as a promising solution owing to its enhanced energy efficiency and seamless integration with prevalent mobile applications. Despite its advantages, the utilization of SoC-Cluster servers remains unsatisfactory, primarily attributed to the tidal patterns of user-initiated workloads. To address such inefficiency, we introduce SoCFlow+, a pioneering framework designed to facilitate the co-location of deep learning training tasks on SoC-Cluster servers, thereby optimizing resource utilization. SoCFlow+ incorporates three novel techniques tailored to mitigate the inherent limitations of commercial SoC-Cluster servers. First, it employs group-wise parallelism complemented by delayed aggregation, a strategy engineered to enhance the training efficiency and scalability of deep learning models, effectively circumventing network bottlenecks. Second, it integrates a data-parallel mixed-precision training algorithm, optimized to exploit the heterogeneous processing capabilities inherent to mobile SoCs fully. Third, SoCFlow+ employs an underclocking-aware workload re-balanacing mechanism to tackle the training performance degradation caused by the thermal control of mobile SoCs. Through rigorous experimental validation, SoCFlow+ achieves a convergence speedup ranging from 1.6x to 740x across 32 SoCs, compared to conventional benchmarks. Furthermore, when juxtaposed with commodity GPU servers (e.g., NVIDIA V100) under identical power constraints, SoCFlow+ not only exhibits comparable training speed but also achieves a remarkable reduction in energy consumption by a factor of 2.31x to 10.23x, all while preserving convergence accuracy.

关键词： Training Servers Task analysis Accuracy Topology Network topology Convergence Edge computing distributed training

来源：评论

学校读者我要写书评

暂无评论

An SDN-Based Flow Table Encoding Approach for Resource and Efficiency Optimization in Topic-Based Pub/Sub Systems

引用

IEEE ACCESS 2024年 12卷 32445-32461页

作者： Zhou, Yu Zhang, Yang Beijing Univ Posts & Telecommun Coll Comp Beijing 100876 Peoples R China Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China

With the rapid development of software-defined networking (SDN), SDN-Based multi-level flow table architectures are always employed to address issues such as QoS, security policies, and matching efficiency. The rise of semantic communication has also sparked researchers' interest in semantic information and its utilization. Many studies on semantic representation and semantic summarization have emerged. This study takes a new perspective to reduce the number of table entries by utilizing the semantic relationships implied in the topic tree in the topic-based pub/sub systems and introduces the concept of semantic aggregation. Semantic aggregation of table entries can work with multi-level flow table architecture to reduce the number of table entries while ensuring the correct delivery of streams. We propose a semantic-based table entry encoding algorithm to implement our idea and conduct several experiments to examine its performance. The experiment results demonstrate that our algorithm can achieve high space and efficiency optimization rate in a short encoding time.

关键词： Semantics Streams Encoding Optimization Security Indexes Aggregates Multilevel systems Software defined networking Semantic aggregation multi-level SDN flow table

来源：评论

学校读者我要写书评

暂无评论

Joint Location Sensing and Demodulation for IRS-Assisted ISAC mmWave MIMO Systems

引用

IEEE TRANSACTIONS ON COMMUNICATIONS 2024年第4期72卷 2470-2484页

作者： Peng, Xingyu Hu, Xiaoling Gan, Xu Zhong, Caijun Zhejiang Univ Coll Informat Sci & Elect Engn Hangzhou Peoples R China Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China

An integrated sensing and communication (ISAC) system assisted by distributed passive intelligent reflecting surfaces (IRSs) is proposed in this paper, which enables simultaneous signal demodulation and location sensing, without channel state information (CSI). A detailed workflow of the proposed IRS-based ISAC system is designed, including transmission protocol, joint location sensing and demodulation, as well as beamforming optimization. Specifically, each coherent block consists of two stages, and each stage is divided into two time blocks, where the signal demodulation, channel estimation, and location sensing are conducted by the proposed integrated localization and demodulation (I-LAD) algorithm, simultaneously. Simulation results show that the signal demodulation performance of the proposed I-LAD algorithm is close to that of the benchmark scheme assuming perfect CSI. At the same time, the channel estimation performance is comparable to that of the parallel factor decomposition (PAPRFAC) method or the compressive sensing-based channel estimation (CS-EST) method. In addition, the proposed location sensing scheme almost achieves the Cramer-Rao lower bound (CRLB) for distributed IRS-assisted localization-only systems, confirming the effectiveness of the proposed I-LAD algorithm.

关键词： Intelligent reflecting surface (IRS) integrated sensing and communication (ISAC) beamforming design location sensing

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：