检索结果-内蒙古大学图书馆

Autonomous Driving via Brain-Inspired Causality-Aware Contrastive Learning with Time-Frequency Prediction

IEEE Internet of Things Journal 2025年

作者： Wang, Chengyu Wang, Luhan Lu, Zhaoming Wu, Celimuge Shou, Guochu Wen, Xiangming Beijing University of Posts and Telecommunications Beijing Key Laboratory of Network System Architecture and Convergence Beijing Laboratory of Advanced Information Networks Beijing China Meta-Networking Research Center The University of Electro-Communications Tokyo182-8585 Japan

Developing trustworthy reinforcement learning (RL) agents for safety-critical control tasks, such as end-to-end autonomous driving, has been a longstanding challenge due to low sample efficiency. Prior works have attempted to address this challenge by performing self-supervised auxiliary tasks like self-reconstruction or predicting long-term future states. However, there still remain unexplored sequential features and causality relationships inherent in sequential state, action, and reward signals in the frequency domain. To fully exploit the temporal and frequential features, we propose a contrastive RL framework called BRain-Inspired causalitY-Aware coNTrastive learning (BRYANT) to achieve efficient representation learning and human-like autonomous driving. Different from existing temporal predictive methods, we transform the sequential latent representations, reward, and action signals into the frequency domain, followed by the symmetric temporal prediction pattern for real and imaginary parts of the frequential signals. To capture the temporal causality for the latent representations, we introduce a brain-inspired network structure called Closed-form Continuous-time (CfC) network to parameterize the derivative of the latent representations and establish the neural dynamic model. Experimental results conducted in the CARLA simulator demonstrate the effectiveness of BRYANT in efficient representation learning, enabling agents to concentrate on potential risks and decrease the collision rate compared to several state-of-the-art RL methods. Furthermore, through the visualization of the latent representation prediction process, we reveal the causal relationships between the critic Q values and the latent representation vectors in the frequency domain, and demonstrate the effectiveness of the frequency domain prediction. © 2014 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Large Scale Model Enabled Semantic Communications Based on Robust Knowledge Distillation

Large Scale Model Enabled Semantic Communications Based on R...

引用

2024 IEEE Global Communications Conference, GLOBECOM 2024

作者： Ding, Kuiyuan Liu, Fangfang Yang, Yang Chen, Mingzhe Guo, Caili Beijing University of Posts and Telecommunications Beijing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing100876 China University of Miami Department of Electrical and Computer Engineering The Institute for Data Science and Computing Coral GablesFL33146 United States Beijing University of Posts and Telecommunications Beijing Laboratory of Advanced Information Networks School of Information and Communication Engineering Beijing100876 China

ISBN: (纸本)9798350351255

Large scale artificial intelligence (AI) models possess excellent capabilities in semantic representation and understanding, making them particularly well-suited for semantic encoding and decoding. However, the substantial scale of these AI models imposes unacceptable computational resources and communication delays. To address this issue, we propose a semantic communication scheme based on robust knowledge distillation (RKD-SC) for large scale model enabled semantic communications. In the considered system, a transmitter extracts the features of the source image for robust transmission and accurate image classification at the receiver. To effectively utilize the superior capability of large scale model while make the cost affordable, we first transfer knowledge from a large scale model to a smaller scale model to serve as the semantic encoder. Then, to enhance the robustness of the system against channel noise, we propose a channel-aware autoencoder (CAA) based on the Transformer architecture. Experimental results show that the encoder of proposed RKD-SC system can achieve over 93.3% of the performance of a large scale model while compressing 96.67% number of parameters. Code: https://***/echojayne/RKD-SC. © 2024 IEEE.

关键词： Channel coding

来源：评论

学校读者我要写书评

暂无评论

Multi-Attention Adaptive Graph Transformer network for Data Inference and Prediction in Vehicular Crowdsensing

Multi-Attention Adaptive Graph Transformer Network for Data ...

引用

International Conference on Intelligent Transportation

作者： Shouyu Huang Luhan Wang Rui Zheng Jie Huo Xiangming Wen Zhaoming Lu Beijing Key Laboratory for Network System Architecture and Convergence Beijing University of Posts and Telecommunications Beijing China School of Traffic and Transportation Beijing Jiaotong University Beijing China

ISBN: (数字)9798331505929

ISBN: (纸本)9798331505936

Vehicular crowdsensing has emerged as an important sensing paradigm in the Internet of Things (IoT), which recruits Intelligent Vehicles to collect data. As its variant, sparse MCS collects data from some sensing areas and infer data for other unsensed areas, addressing the challenge of maintaining high-quality sensing at manageable costs. In practical vehicular crowdsensing scenarios, the diverse requirements of urban applications make it crucial not only to infer data for the current period but also to predict the entire sensing map for the future. This predictive capability not only reduces the cost of sensing but also provides critical data support for various urban applications, including intelligent transportation systems. This paper introduces the Multi-Attention Adaptive Graph Transformer network (MAAGTN) aimed at improving the accuracy of data inference and prediction in vehicular crowdsensing. We present a novel embedding method that integrates spatial-temporal information of sensing maps into the model. An adaptive learnable graph module and a multiattention module are designed to dynamically capture complex spatial-temporal correlations among data. Furthermore, a dynamic multi-task learning framework is proposed to mitigate error propagation by adjusting task weights during training. Experimental evaluations on real-world datasets demonstrate the superiority of MAAGTN for data inference and prediction.

关键词： Training Adaptive systems Costs Correlation Crowdsensing Transformers Multitasking Sensors Internet of Things Vehicle dynamics

来源：评论

学校读者我要写书评

暂无评论

Signal-SGN: A Spiking Graph Convolutional network for Skeleton Action Recognition via Learning Temporal-Frequency Dynamics

arXiv

引用

arXiv 2024年

作者： Zheng, Naichuan Du, Yuchen Xia, Hailun Liu, Dapeng Beijing Laboratory of Advanced Information Networks Beijing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing100876 China

For skeleton-based action recognition, Graph Convolutional networks (GCNs) are effective models. Still, their reliance on floating-point computations leads to high energy consumption, limiting their applicability in battery-powered devices. While energy-efficient, Spiking Neural networks (SNNs) struggle to model skeleton dynamics, leading to suboptimal solutions. We propose Signal-SGN (Spiking Graph Convolutional network), which utilizes the temporal dimension of skeleton sequences as the spike time steps and represents features as multi-dimensional discrete stochastic signals for temporal-frequency domain feature extraction. It combines the 1D Spiking Graph Convolution (1D-SGC) module and the Frequency Spiking Convolution (FSC) module to extract features from the skeleton represented as spiking form. Additionally, the Multi-Scale Wavelet Transform Feature Fusion (MWTF) module is proposed to extract dynamic spiking features and capture frequency-specific characteristics, enhancing classification performance. Experiments across three large-scale datasets reveal Signal-SGN exceeding state-of-the-art SNN-based methods in accuracy and computational efficiency while attaining comparable performance with GCN methods and significantly reducing theoretical energy consumption. Copyright © 2024, The Authors. All rights reserved.

关键词： Musculoskeletal system

来源：评论

学校读者我要写书评

暂无评论

Differential Electrical Conductivity Strength Based Neighborhood Compactness Model: A Novel Vulnerable Node Identification Method for Power networks

SSRN

引用

SSRN 2024年

作者： Min, Zhao Junhan, Ye Jiayun, Li Beijing Laboratory of Advanced Information Networks Beijing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing100876 China

The large scale and complex connections of the power networks make the vulnerability of the networks increasingly prominent. Due to the sensitivity of networks to interference, the failure of a single vulnerable node can quickly spread and bring down the system. Therefore, it is of great significance to efficiently and accurately identify the vulnerable nodes to improve the safety and management of the power networks. The traditional global information-based vulnerable node identification methods have high computational complexity and cannot meet the real-time requirements of the power networks. The existing methods based on local information only focus on the information of neighbors but lack the exploration of the interaction between neighbors. This paper focuses on the essential characteristics of the power network topology and attempts to mine the vulnerable nodes through broader and richer neighborhood information. Further, a novel vulnerable node identification method combining the neighborhood compactness model and differential electrical conductivity strength (ECS) is proposed. From the new perspective of the decentered neighborhood, we propose the kth-order neighborhood ECS to quantitatively analyze the strength of the multipath connections of the neighbors. Meanwhile, this paper defines the differential ECS to describe the closeness of the neighbors considering the distance between the central node and the paths between neighbors. And finally, the identification of vulnerable nodes is achieved by the neighborhood compactness model. Our newly proposed parameter, average ECS, as well as the traditional parameters, including network efficiency, connectivity, vulnerability, capacity, ranking distribution, and complexity, are used for the performance analysis of the algorithm. The proposed algorithm generally outperforms the benchmark methods in the above six dimensions in our comparison experiments on nine power networks. © 2024, The Authors. All rights reserv

关键词： network topology

来源：评论

学校读者我要写书评

暂无评论

Design of a Low-Cost Compact Antenna with 360° Beam Scanning 13

Design of a Low-Cost Compact Antenna with 360° Beam Scannin...

引用

13th International Conference on Microwave and Millimeter Wave Technology, ICMMT 2021

作者： Liu, Tao Chen, Yuxing Hong, Weijun Beijing University Of Posts And Telecommunications Beijing Key Laboratory Of Network System Architecture And Convergence Beijing Laboratory Of Advanced Information Network Beijing100876 China

ISBN: (纸本)9781665434379

In this paper, a low-cost, compact antenna system for wideband spectrum sensing and 360° beam scanning is designed. The system consists of an elliptical monopole Ultra-Wideband (UWB) antenna, and a cylindric antenna radome is composed of a PIN diode array and copper patch array. The passband/stopband reconfigurable Frequency selective surface(FSS) radome model is established by switching the state of the PIN diode array, and a 360° beam scanning switch is realized. The simulation results show that the operation band is 3.5GHz-11.8GHz, and the relative bandwidth is up to 108% in the wideband working mode. In the beam scanning mode, the -3dB main lobe width is about 35°-48°, achieving the 360° beam scanning with 10° as the step. © 2021 IEEE.

关键词： Frequency selective surfaces

来源：评论

学校读者我要写书评

暂无评论

MK-SGN: A Spiking Graph Convolutional network with Multimodal Fusion and Knowledge Distillation for Skeleton-based Action Recognition

arXiv

引用

arXiv 2024年

作者： Zheng, Naichuan Xia, Hailun Liang, Zeyu Du, Yuchen Beijing Laboratory of Advanced Information Networks Beijing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing100876 China

In recent years, multimodal Graph Convolutional networks (GCNs) have achieved remarkable performance in skeleton-based action recognition. The reliance on high-energy-consuming continuous floating-point operations inherent in GCN-based methods poses significant challenges for deployment in energy-constrained, battery-powered edge devices. To address these limitations, MK-SGN, a Spiking Graph Convolutional network with Multimodal Fusion and Knowledge Distillation, is proposed to leverage the energy efficiency of Spiking Neural networks (SNNs) for skeleton-based action recognition for the first time. By integrating the energy-saving properties of SNNs with the graph representation capabilities of GCNs, MK-SGN achieves significant reductions in energy consumption while maintaining competitive recognition accuracy. Firstly, we formulate a Spiking Multimodal Fusion (SMF) module to effectively fuse multimodal skeleton data represented as spike-form features. Secondly, we propose the Self-Attention Spiking Graph Convolution (SA-SGC) module and the Spiking Temporal Convolution (STC) module, to capture spatial relationships and temporal dynamics of spike-form features. Finally, we propose an integrated knowledge distillation strategy to transfer information from the multimodal GCN to the SGN, incorporating both intermediate-layer distillation and soft-label distillation to enhance the performance of the SGN. MK-SGN exhibits substantial advantages, surpassing state-of-the-art GCN frameworks in energy efficiency and outperforming state-of-the-art SNN frameworks in recognition accuracy. The proposed method achieves a remarkable reduction in energy consumption, exceeding 98% compared to conventional GCN-based approaches. This research establishes a robust baseline for developing high-performance, energy-efficient SNN-based models for skeleton-based action recognition Copyright © 2024, The Authors. All rights reserved.

关键词： Musculoskeletal system

来源：评论

学校读者我要写书评

暂无评论

Joint Optimization of Base Station Sleeping, Functional Split, and Routing Selection in Virtualized Radio Access networks

Joint Optimization of Base Station Sleeping, Functional Spli...

引用

IEEE Conference on Wireless Communications and networking

作者： Yunqi Xu Hang Li Zhenghe Zhu Yawen Chen Luhan Wang Zhaoming Lu Xiangming Wen The Beijing Key Laboratory of Network System Architecture and Convergence Beijing Laboratory of Advanced Information Networks Beijing University of Posts and Telecommunications Beijing China

This paper investigates the Joint Optimization of Base station sleeping, Functional split, and Routing selection (JOBFR) in virtualized radio access network (vRAN) architectures to minimize the operator’s total cost, including operating cost and migration cost while satisfying user requirements. Specifically, we first use mathematical methods to describe the relationships between the base station (BS) sleeping, functional split (FS) and routing selection, and then model the goal as a joint optimization problem. Next, we propose a heuristic algorithm called Flexible Sleeping, Functional split, and Routing selection (FSFR) to decide BS sleeping, and FS option, and routing selection, simultaneously. Finally, we perform extensive simulations to evaluate our algorithm. The results demonstrate that our algorithm can significantly save the total cost of operators compared to the baseline approach and can also achieve near-optimal performance.

关键词： Base stations Costs Heuristic algorithms Routing Mathematical models Optimization Radio access networks

来源：评论

学校读者我要写书评

暂无评论

Super-Resolution Enhanced Video Streaming Based on Cloud-Edge Cooperation for Satellite Backhauls

Super-Resolution Enhanced Video Streaming Based on Cloud-Edg...

引用

IEEE Globecom Workshops

作者： Changhao Liu Wenpeng Jing Ziyuan Zheng Zhaoming Lu Xiangming Wen Beijing Key Laboratory of Network System Architecture and Convergence Beijing Laboratory of Advanced Information Networks Beijing University of Posts and Telecommunications Beijing China

Although satellite backhauls facilitate expanding network coverage and providing capacity in unconnected areas, delivering high-quality videos via satellite backhauls is challenging due to high delay, bandwidth limitations and high expenses. Super-resolution (SR) technique has been proposed to integrate into streaming systems to improve Quality of Experience (QoE) in bandwidth-limited environments. However, existing SR-enhanced streaming systems face two key challenges: 1) compatibility problem that legacy video servers and user equipment (UEs) must be modified to integrate SR-related modules, and 2) quality consistency problem that enhanced video quality fluctuates across chunks. Motivated by this, we present SatOL, a novel SR-enhanced video streaming system for satellite-backhauled networks. Specifically, SatOL is designed on a cloud-edge cooperation framework, where a cloud server and an edge server are responsible for SR model training and SR-based video enhancement, respectively. This framework operates as an intermediate hop within end-to-end streaming sessions to perform SR-related processing, avoiding modifications to legacy video servers and UEs. Instead of training video-tailored SR models offline, the cloud server continually trains chunk-tailored SR models during video streaming, which are overfitted to corresponding chunks and provide consistent quality enhancement. To avoid rebuffering caused by time-consuming model training, an online fine-tuning method is proposed to significantly expedite SR model training. To further improve online training efficiency, we propose a decision mechanism that determines whether and how to train chunk-tailored SR models based on the chunks' 1-frame count. Evaluations show that SatOL outperforms baselines, improving QoE and enhanced video quality by 6-285% and 17-25%, respectively, while reducing bandwidth usage by 60-64%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Mobility-Aware AR Computation Tasks Offloading Based on Offline RL over Terahertz system

Mobility-Aware AR Computation Tasks Offloading Based on Offl...

引用

IEEE Globecom Workshops

作者： Shuyue Zhao Wenpeng Jing Xiangming Wer Zhaoming Lu Beijing Key Laboratory of Network System Architecture and Convergence Beijing Laboratory of Advanced Information Networks Beijing University of Posts and Telecommunications Beijing China

Augmented reality (AR) technology is increasingly applied in the Internet of vehicles (IoV). As it poses significant challenges to the communication and computation capabilities of AR terminal devices, this paper investigates combining terahertz (THz) and mobile edge computing (MEC) integrated with reinforcement learning (RL) to optimize the performance of AR services for the IoV. Though the RL-based MEC offloading has been validated to have excellent performance, it has an intrinsic disadvantage where its training phase needs inappropriate offloading strategies to be tested frequently and poses risks to safe driving. Moreover, although THz communication can provide high transmission data rates, its characteristics that different transmission distances lead to different available bandwidth, pose a unique challenge for the communication resources allocation. Motivated by these, we propose a novel static data-driven AR tasks offloading framework to solve resources allocation problem for the IoV with THz communication. To deal with the challenge that different users have different available THz bandwidth, we propose a distance-aware fair THz bandwidth allocation algorithm. Moreover, the offline RL, instead of online RL, is utilized to train the AR task offloading decision agent, where interaction with the real world is not required, completely safeguarding the system's safety. To address the typical issue of out-of-distribution (ODD) actions encountered in offline RL, we adopt conservative Q-learning (CQL) to train and obtain the optimal AR computation task offloading strategy to enhance users' experience. Simulation results show that the proposed solution outperforms state-of-the-art algorithms in terms of algorithm stability and user reward.

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：