检索结果-内蒙古大学图书馆

arXiv 2019年

作者： Liu, Xuanlin Chen, Mingzhe Yin, Changchuan Saad, Walid Beijing Key Laboratory of Network System Architecture and Convergence Beijing University of Posts and Telecommunications Beijing China Wireless@VT Bradley Department of Electrical and Computer Engineering Virginia Tech BlacksburgVA United States

—In this paper, the echo state network (ESN) memory capacity, which represents the amount of input data an ESN can store, is analyzed for a new type of deep ESNs. In particular, two deep ESN architectures are studied. First, a parallel deep ESN is proposed in which multiple reservoirs are connected in parallel allowing them to average outputs of multiple ESNs, thus decreasing the prediction error. Then, a series architecture ESN is proposed in which ESN reservoirs are placed in cascade that the output of each ESN is the input of the next ESN in the series. This series ESN architecture can capture more features between the input sequence and the output sequence thus improving the overall prediction accuracy. Fundamental analysis shows that the memory capacity of parallel ESNs is equivalent to that of a traditional shallow ESN, while the memory capacity of series ESNs is smaller than that of a traditional shallow ESN. In terms of normalized root mean square error, simulation results show that the parallel deep ESN achieves 38.5% reduction compared to the traditional shallow ESN while the series deep ESN achieves 16.8% reduction. Copyright © 2019, The Authors. All rights reserved.

关键词： network architecture

来源：评论

学校读者我要写书评

暂无评论

Multi-Agent Deep Reinforcement Learning based Spectrum Allocation for D2D Underlay Communications

arXiv

引用

arXiv 2019年

作者： Li, Zheng Guo, Caili School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing100876 China Beijing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing100876 China

—Device-to-device (D2D) communication underlay cellular networks is a promising technique to improve spectrum efficiency. In this situation, D2D transmission may cause severe interference to both the cellular and other D2D links, which imposes a great technical challenge to spectrum allocation. Existing centralized schemes require global information, which causes a large signaling overhead. While existing distributed schemes requires frequent information exchange among D2D users and cannot achieve global optimization. In this paper, a distributed spectrum allocation framework based on multi-agent deep reinforcement learning is proposed, named multi-agent actor critic (MAAC). MAAC shares global historical states, actions and policies during centralized training, requires no signal interaction during execution and utilizes cooperation among users to further optimize system performance. Moreover, in order to decrease the computing complexity of the training, we further propose the neighbor-agent actor critic (NAAC) based on the neighbor users’ historical information for centralized training. The simulation results show that the proposed MAAC and NAAC can effectively reduce the outage probability of cellular links, greatly improve the sum rate of D2D links and converge quickly. Copyright © 2019, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Federated learning for task and resource allocation in wireless high altitude balloon networks

arXiv

引用

arXiv 2020年

作者： Wang, Sihua Chen, Mingzhe Yin, Changchuan Saad, Walid Hong, Choong Seon Cui, Shuguang Vincent Poor, H. Beijing Laboratory of Advanced Information Network Beijing Key Laboratory of Network System Architecture and Convergence Beijing University of Posts and Telecommunications Beijing100876 China Department of Electrical Engineering Princeton University PrincetonNJ08544 United States Wireless@VT Bradley Department of Electrical and Computer Engineering Virginia Tech BlacksburgVA24060 United States Department of Computer Science and Engineering Kyung Hee University Korea Republic of Chinese University of Hong Kong Shenzhen518172 China Department of Electrical Engineering Princeton University PrincetonNJ08544 United States

In this paper, the problem of minimizing energy and time consumption for task computation and transmission is studied in a mobile edge computing (MEC)-enabled balloon network. In the considered network, each user needs to process a computational task in each time instant, where high-altitude balloons (HABs), acting as flying wireless base stations, can use their powerful computational abilities to process the tasks offloaded from their associated users. Since the data size of each user's computational task varies over time, the HABs must dynamically adjust the user association, service sequence, and task partition scheme to meet the users' needs. This problem is posed as an optimization problem whose goal is to minimize the energy and time consumption for task computing and transmission by adjusting the user association, service sequence, and task allocation scheme. To solve this problem, a support vector machine (SVM)-based federated learning (FL) algorithm is proposed to determine the user association proactively. The proposed SVM-based FL method enables each HAB to cooperatively build an SVM model that can determine all user associations without any transmissions of either user historical associations or computational tasks to other HABs. Given the prediction of the optimal user association, the service sequence and task allocation of each user can be optimized so as to minimize the weighted sum of the energy and time consumption. Simulations with real data of city cellular traffic from the OMNILab at Shanghai Jiao Tong University show that the proposed algorithm can reduce the weighted sum of the energy and time consumption of all users by up to 16.1% compared to a conventional centralized method. Copyright © 2020, The Authors. All rights reserved.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

A visualization compression algorithm considering interactive structure for energy trading of microgrid system 2

A visualization compression algorithm considering interactiv...

引用

2nd IEEE International Conference on Energy Internet, ICEI 2018

作者： Shen, Yidi Wen, Xiangming Wang, Luhan Guan, Wanqing Beijing Advanced Innovation Center for Future Internet Technology Beijing China Beijing Laboratory of Advanced Information Networks Beijing University of Posts and Telecommunications Beijing China Beijing Key Laboratory of Network System Architecture and Convergence Beijing China

ISBN: (纸本)9781538641316

A microgrid is a novel energy production-storageconsumption mode in which energy entities can both generate and consume energy. With the access of these entities, the energy internet becomes a super complex network of information and physics integration. The computational difficulty of obtaining the topological information is improved. Studying the topology for energy trading of energy entities in different geographic regions can help to analyze the transmission of energy and information. Visualization is an important measure to display the network topological structure intuitively and clearly. In this paper, a compression algorithm that considered cross-regional interaction is proposed to obtain the main topology for visualization effectively. The proposed algorithm can realize the different degrees of compression between intra-regional topology and inter-regional topology. The simulation results have shown that the algorithm can preserve the main topological information and has a considerable compression rate. © 2018 IEEE.

关键词： Visualization

来源：评论

学校读者我要写书评

暂无评论

Dynamic Resource Orchestration of Service Function Chaining in network Function Virtualizations 1st

Dynamic Resource Orchestration of Service Function Chaining ...

引用

1st International Conference on 5G for Future Wireless networks, 5GWN 2017

作者： Yu, Bangchao Zheng, Wei Wen, Xiangming Lu, Zhaoming Wang, Luhan Ma, Lu Beijing Laboratory of Advanced Information Networks Beijing China Beijing Advanced Innovation Center for Future Internet Technology Beijing China Beijing Key Laboratory of Network System Architecture and Convergence Beijing University of Posts and Telecommunications Beijing China

ISBN: (纸本)9783319728223

network Functions Virtualization is a new network architecture framework and is revolutionizing the way networking service that how to design and deploy. NFV promotes virtualizing network functions and improves the flexibility to resource orchestration for request service function chains. However, how to find the most suitable resource in NFV-based network resource is a challenge. This paper presents a comprehensive state of the NFV resource orchestration by introducing a dynamic resource orchestration architecture that can configure dynamic resources. With consideration of load balance, energy cost and resource consumption, the resource orchestration is formulated as a multi-objective optimal problem. Finally, a multi-objective simulated annealing algorithm is used to obtain the optimal resource strategy to deploy network functions. Simulation results show that the solution for dynamic resource orchestration can achieve approximate optimal solution in acceptable time and reduce 8% energy consumption with a 0.89 Jain’s fairness index. © 2018, ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering.

关键词： network function virtualization

来源：评论

学校读者我要写书评

暂无评论

A fault-tolerant data acquisition scheme with MDS and dynamic clustering in energy internet 2

A fault-tolerant data acquisition scheme with MDS and dynami...

引用

2nd IEEE International Conference on Energy Internet, ICEI 2018

作者： Chen, Xin Wen, Xiangming Wang, Luhan Jing, Wenpeng Beijing Laboratory of Advanced Information Networks Beijing University of Posts and Telecommunications Beijing China Beijing Key Laboratory of Network System Architecture and Convergence Beijing China Beijing Advanced Innovation Center for Future Internet Technology Beijing China

ISBN: (纸本)9781538641316

Energy Internet (EI) is regarded as a promising method to solve the energy shortage and improve the efficiency of renewable energy. It demands frequent exchanges of data flow and energy flow, which puts forward strict requests for data veracity and availability. As a result, a reliable and robust data network is critical to realize the promises of EI. This paper introduces the fog nodes with computing ability to enable erasure coding based fault-tolerant techniques so that the data availability can be guaranteed. However, the distributed and irregular deployment of EI nodes have brought many challenges to maximum distance separable code (MDS) based fault-tolerant schemes. As a result, this paper investigates the design of encoding parameters and clustering mechanism to adjust the flexible situations. And then, we propose a novel fault-tolerant mechanism based on MDS and dynamic clustering to improve the data availability. Numerical simulation results show our scheme works better than the backup and traditional MDS based schemes with determined cluster size. © 2018 IEEE.

关键词： Internet of Energy

来源：评论

学校读者我要写书评

暂无评论

Exploiting offset-guided network for pose estimation and tracking

arXiv

引用

arXiv 2019年

作者： Zhang, Rui Zhu, Zheng Li, Peng Wu, Rui Guo, Chaoxu Huang, Guan Xia, Hailun Bejing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing China Beijing Laboratory of Advanced Information Networks Beijing Institute of Automation Chinese Academy of Sciences Beijing China Horizon Robotics Beijing

Human pose estimation has witnessed a significant advance thanks to the development of deep learning. Recent human pose estimation approaches tend to directly predict the location heatmaps, which causes quantization errors and inevitably deteriorates the performance within the reduced network output. Aim at solving it, we revisit the heatmap-offset aggregation method and propose the Offset-guided network (OGN) with an intuitive but effective fusion strategy for both two-stages pose estimation and Mask R-CNN. For two-stages pose estimation, a greedy box generation strategy is also proposed to keep more necessary candidates while performing person detection. For mask R-CNN, ratio-consistent is adopted to improve the generalization ability of the network. State-of-the-art results on COCO and PoseTrack dataset verify the effectiveness of our offset-guided pose estimation and tracking. Copyright © 2019, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Ultra-broadband Polarization Rotator Design in Terahertz Region

Ultra-broadband Polarization Rotator Design in Terahertz Reg...

引用

International Conference on Wireless Communications and Signal Processing (WCSP)

作者： Weijun Hong Xinlong Li Jianfeng Zhu Shufang Li Tingxiao Cai Huanhuan Lin Hongjie Liu Yanqing Wei Beijing Key Laboratory of Network System Architecture and Convergence Beijing University of Posts and Telecommunications Beijing China Beijing Laboratory of Advanced Information Network Beijing University of Posts and Telecommunications Beijing China China Electronics Standardization Institute Beijing China Borsche Technologies Limited Company Beijing China Renmin University of China Beijing China

A three layered metamaterial functions as polarization rotator with extremely broad bandwidth and high conversion ratio is proposed and investigated in this paper. The designed structure is composed of two electric dipole resonant polarization selectors on the top and bottom sides and a cross-shaped polarization converter in the middle filled with two dielectric layers of silica as spacers. The proposed structure forms a cascaded Fabry-Perot-like resonance cavity, leading to the enhancement of polarization conversion between two orthogonal linearly polarized waves. The designed structure is optimized to achieve an extremely broad bandwidth, sharp rejection stop-band, and maximum conversion ratio of 99.9% at the resonant frequency.

关键词： Metamaterials Cavity resonators Polarization Bandwidth Substrates Telecommunications Dielectrics

来源：评论

学校读者我要写书评

暂无评论

Energy Consumption Optimization Scheme of Cloud Data Center Based on SDN

引用

Procedia Computer Science 2018年 131卷 1318-1327页

作者： Qing Liao Zun Wang Beijing Key Laboratory of Network System Architecture and Convergence Beijing University of Posts and Telecommunications Beijing 100876 China

In response to the mobile Internet, Internet of things and cloud computing technology fast development, the operation of large-scale distributed computing cloud data center construction in the world at the same time, the network data calculation of energy consumption problem is causing extensive concern. And the software-defined network (SDN) provides a feasible solution to the rigid problem of traditional network architecture. Based on SDN technology, this paper proposes a method of energy consumption control based on cloud data center. Aiming at server energy saving, the multi-objective optimization model of virtual machine migration integration is established, and the integration mechanism of virtual machine migration based on mixed single-parent genetic algorithm is proposed. The experimental results show that the algorithm can be used to solve the problem of virtual machine migration and integration in data center.

关键词： Cloud data center software-defined network energy consumption control VM migration integration

来源：评论

学校读者我要写书评

暂无评论

Generative Model with Kernel Density Estimation

Generative Model with Kernel Density Estimation

引用

IEEE International Conference on network Infrastructure and Digital Content (IC-NIDC)

作者： Jing Tan Sixing Yin Shuo Zhao Beijing Key Laboratory of Network System Architecture and Convergence Beijing China

In this paper, a novel generative model is proposed based on kernel density estimation. Different from other generative models like GANs [1] and VAEs [2], the model directly learns the distribution of training dataset in a non-parametric way. The mean absolute error (MAE) between the estimated probability density of the real and generated samples is used as loss function for model to train. To address the sparsity of high-dimensional distribution, we associate the proposed generative model with an encoder, by which the observed high-dimensional data can be transformed to low-dimensional latent features. Results on the MNIST dataset show that our proposed model is able to generate samples visually indistinguishable from the real ones.

关键词： Kernel Estimation Training Generators Training data Gallium nitride Bandwidth

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：