检索结果-南通市图书馆

Semantic separator learning and its applications in unsupervised Chinese text parsing

在线全文

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2013年第1期7卷 55-68页

作者： Yuming WU Xiaodong LUO Zhen YANG Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing 100190 China Graduate University of the Chinese Academy of Sciences Beijing 100049 China china telecom corporation limited Shanghai Branch Shanghai 200120 China Shanghai research Institute of china telecom corporation limited Shanghai 200120 China

Grammar learning has been a bottleneck problem for a long time. In this paper, we propose a method of seman- tic separator learning, a special case of grammar learning. The method is based on the hypothesis that some classes of words, called semantic separators, split a sentence into sev- eral constituents. The semantic separators are represented by words together with their part-of-speech tags and other infor- mation so that rich semantic information can be involved. In the method, we first identify the semantic separators with the help of noun phrase boundaries, called subseparators. Next, the argument classes of the separators are learned from cor- pus by generalizing argument instances in a hypernym space. Finally, in order to evaluate the learned semantic separators, we use them in unsupervised Chinese text parsing. The exper- iments on a manually labeled test set show that the proposed method outperforms previous methods of unsupervised text parsing.

关键词： semantic separator separator learning unsuper vised text parsing

Group-based joint signaling and data resource allocation in MTC-underlaid cellular networks

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science china(Information Sciences) 2017年第10期60卷 55-67页

作者： Xuefei ZHANG Yue WANG Ruiqi WANG Yanzhao HOU Da WANG School of Information and Communication Engineering Beijing University of Posts and Telecommunications china telecom corporation limited Technology Innovation Center

Machine-type communications(MTC) are gaining significant research attention as one of the most promising technologies for the fifth generation(5G) mobile networks. A critical issue handled by MTC is support for massive numbers of connections, which is a growing problem that will become increasingly challenging as MTC share spectrum resources with cellular communication. Here, not only the number of connections but also the data rate requirements of cellular users(CUEs) need to be considered. Given these issues, in this paper, we formulate a group-based joint signaling and data resource optimization model constrained by network resource and data rate requirements in order to maximize the number of connections. We also note that this problem is nonconvex and that obtaining an optimal solution is computationally complex for MTC with massive numbers of users(UEs). Therefore, we decompose the problem into group-based data aggregation and resource allocation subproblems. To solve these two subproblems, we develop an adaptive group head selection algorithm and a joint signaling and data resource allocation algorithm that satisfy both the data rate requirement and resource constraints, respectively. Our simulation results show that our proposed algorithms significantly improve the number of connections when compared with other classic methods. Furthermore, our results reveal that the limiting factor on the number of connections changes with the ratio of the number of MTC UEs to that of CUEs and the ratio of data requirement of MTC UEs to that of CUEs. Finally, we note that our proposed group-based resource allocation algorithm can effectively improve the number of connections, especially when more MTC UEs and a small amount of MTC data are present.

关键词： data aggregation matrix analysis machine-type communications signaling and data resource allocation

Integrated Sensing and Communication Enabled Multiple Beamwidth and Power Allocation for Connected Automated Vehicles

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

中国通信:英文版 2023年第9期20卷 46-58页

作者： Shengnan Liu Qianyi Hao Qixun Zhang Jiaxiang Liu Zheng Jiang china telecom corporation limited Beijing research Institute Beijing 102209China School of Information and Communication Engineering Beijing University of Posts and TelecommunicationsBeijing 100876China

Connected autonomous vehicles(CAVs)are a promising paradigm for implementing intelligent transportation ***,in CAVs scenarios,the sensing blind areas cause serious safety *** vehicle-to-vehicle(V2V)technology is difficult to break through the sensing blind area and ensure reliable sensing *** overcome these problems,considering infrastructures as a means to extend the sensing range is feasible based on the integrated sensing and communication(ISAC)*** mmWave base station(mmBS)transmits multiple beams consisting of communication beams and sensing *** sensing beams are responsible for sensing objects within the CAVs blind area,while the communication beams are responsible for transmitting the sensed information to the *** reduce the impact of inter-beam interference,a joint multiple beamwidth and power allocation(JMBPA)algorithm is *** maximizing the communication transmission rate under the sensing *** proposed non-convex optimization problem is transformed into a standard difference of two convex functions(D.C.)***,the superiority of the *** average transmission rate of communication beams remains over 3.4 Gbps,showcasing a significant improvement compared to other ***,the satisfaction of sensing services remains steady.

关键词： integrated sensing and communication connected automated vehicles resource allocation in wireless networks

Intelligent reflecting surface-assisted federated learning in multi-platoon collaborative networks

在线全文

学校读者我要写书评

暂无评论

Digital Communications and Networks 2023年第3期9卷 628-637页

作者： Xiaoting Ma Junhui Zhao Jieyu Liao Ziyang Zhang School of Electronic and Information Engineering Beijing Jiaotong UniversityBeijing100044China china telecom corporation limited research Institute Beijing102209China School of Information Engineering East China Jiaotong UniversityNanchang330013China

Inspired by mobile edge computing(MEC),edge learning has gained a momentum by directly performing model training at network edge without sending massive data to a centralized data ***,the quality of model training will be affected by the limited communication and computing resources of network *** this paper,how to improve the training performance of a federated learning system aided by intelligent reflecting surface(IRS)over vehicle platooning networks is studied,where multiple platoons train a shared federated learning ***-platoon cooperation can alleviate the pressure of data processing caused by the limited computing resources of single ***,IRS can enhance the inter-platoon communication in a cost-effective and energy-efficient ***,the federated learning optimization problem of maximizing the learning accuracy is formulated by jointing platoon scheduling,bandwidth allocation and phase shifts at the IRS to maximize the number of scheduled ***-ically,in the proposed learning architecture each platoon updates the learning model with its own data and uploads it to the global model through IRS-based wireless ***,a method based on sequential optimization algorithm(SOA)and a group-based optimization method are analyzed for single IRS aided and large-scale IRS aided commu-nication,***,a platoon scheduling scheme is designed based on the communication reliability and computing reliability of *** results demonstrate that large-scale IRS assisted communication can effectively improve the reliability of multi-user communication *** scheduling scheme based on learning reliability balances the communication performance and computing performance of platoons.

关键词： Vehicle platooning networks Federated learning Intelligent reflecting surface Platoon scheduling Resource allocation

Architecture for Cellular Enabled Integrated Communication and Sensing Services

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

中国通信:英文版 2023年第9期20卷 59-77页

作者： Bo Liu Qixun Zhang Zheng Jiang Dongsheng Xue Chenlong Xu Bowen Wang Xiaoming She Jinlin Peng School of Information and Communication Engineering Beijing University of Posts and TelecommunicationsBeijing 100876China china telecom corporation limited Beijing research Institute Beijing 102209China Artificial Intelligence research Center National Innovation Institute of Defense TechnologyBeijingChina

There is growing interest in the integrated sensing and communication(ISAC)to extend the 5G+/6G network capabilities by introducing sensing *** the solutions for mono-static or bi-static ISAC have shown feasibility and benefits based on existing 5G physical layer design,whether and how to coordinate multiple ISAC devices to better exert networking performance are rarely discussed.3 rd Partnership Project(3GPP)has initiated the ISAC use cases study,and the follow-up studies for network architecture could be *** this article,we focus on gNB-based sensing mode and propose ISAC functional framework with given of highlevel service procedures to enable cellular based ISAC *** the proposed ISAC framework,three types of network functions for sensing service as Sensing Function(SF),lightweight-Edge Sensing Function(ESF)and full-version-ESF are designed with interaction with network nodes to fulfill the latency requirements of ISAC use ***,with simulation evaluations and hardware testbed results,we further verify the performance benefit and feasibility to enable ISAC in 5G for the gNB-based sensing mode with new design on SF and related signaling protocols.

关键词： integrated sensing and communication(ISAC) network architecture 5G system

Multi-pair massive MIMO amplify-and-forward relaying system with low-resolution ADCs:performance analysis and power control

在线全文

学校读者我要写书评

暂无评论

Science china(Information Sciences) 2018年第2期61卷 164-181页

作者： Qimei CUI Yinjun LIU Yinsheng LIU Weiliang XIE Yong ZHAO National Engineering Laboratory for Mobile Network Security Beijing University of Posts and Telecommunications State Key Laboratory of Rail Traffic Control and Safety Beijing Jiaotong University china telecom corporation limited

In this paper, we focus on a general multi-pair massive MIMO amplify-and-forward(AF) relaying system where the relay antennas employ low-resolution analog-to-digital converters(ADCs) to reduce the hardware cost. First, considering the effect of low quantization on channel estimation, a tight closed form approximation of the system ergodic achievable rate is derived. Second, some asymptotic analysis is presented to reveal the impacts of the system parameters on the achievable rate. Particularly, the generalized power scaling schemes are characterized. The results indicate that in some cases, when the number of relay antennas grows without bound, the impact of the finite resolution ADCs on data transmission can be eliminated. To enhance the achievable rate of the quantized systems, the optimal user and relay power control schemes are proposed. Furthermore, to reap all the benefits of low-resolution ADCs, another power control scheme is also designed to minimize the total power consumption while guaranteeing the quality-of-service(Qo S)requirement of each user, which can help draw some useful insights into the optimal ADC resolution from power saving perspectives. The simulation results confirm the accuracy of our theoretical analysis and the effectiveness of the proposed power control schemes.

关键词： massive MIMO relaying system imperfect CSI low quantization power control

Cognitive RAN Slicing Resource Allocation Based on Stackelberg Game

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

china Communications 2022年第5期19卷 12-23页

作者： Tengteng Ma Yong Zhang Siyu Yuan Zhenjie Cheng School of Electronic Engineering Beijing University of Posts and TelecommunicationsBeijing 100876China Beijing Key Laboratory of Work Safety Intelligent Monitoring(Beijing University of Posts and telecommunications) Beijing 100876China Network Planning research and innovation Center China Telecom Corporation Limited Research InstituteBeijing 102209China

The cognitive network has become a promising method to solve the spectrum resources shortage *** for the optimization of network slicing resources in the cognitive radio access network(RAN),we are interested in the profit of the mobile virtual network operator(MVNO)and the utility of secondary users(SUs).In cognitive RAN,we aim to find the optimal scheme for the MVNO to efficiently allocate slice resources to *** the MVNO and SUs are selfish and the game between the MVNO and SUs is difficult to reach equilibrium,we consider modeling this scheme as a Stackelberg *** mathematical programming with equilibrium constraints(MPEC)and Karush-Kuhn-Tucker(KKT)conditions,we can obtain a single-level optimization problem,and then prove that the problem is a convex optimization *** simulation results show that the proposed method is superior to the noncooperative *** guaranteeing the Quality of Service(QoS)of primary users(PUs)and SUs,the proposed method can balance the profit of the MVNO and the utility of SUs.

关键词： cognitive RAN network slicing MVNO Stackelberg game MPEC KKT

User Scheduling and Slicing Resource Allocation in Industrial Internet of Things

在线全文

学校读者我要写书评

暂无评论

china Communications 2023年第6期20卷 368-381页

作者： Sisi Li Yong Zhang Siyu Yuan Tengteng Ma Beijing University of Posts and telecommunications Beijing 100876China Beijing Key Laboratory of Work Safety Intelligent Monitoring Beijing University of Posts and TelecommunicationsBeijing 100876China Strategy Development Institute China Telecom Corporation Limited Beijing Research InstituteBeijing 102209China

Heterogeneous base station deployment enables to provide high capacity and wide area *** slicing makes it possible to allocate wireless resource for heterogeneous services on *** two promising technologies contribute to the unprecedented service in *** establish a multiservice heterogeneous network model,which aims to raise the transmission rate under the delay constraints for active control terminals,and optimize the energy efficiency for passive network terminals.A policygradient-based deep reinforcement learning algorithm is proposed to make decisions on user association and power control in the continuous action *** results indicate the good convergence of the algorithm,and higher reward is obtained compared with other baselines.

关键词： wireless communication resource allocation reinforcement learning heterogeneous network network slicing Internet of Things

Workpiece Surface Defect Detection Based on Prototype Network With Blur Pooling

在线全文

学校读者我要写书评

暂无评论

Workpiece Surface Defect Detection Based on Prototype Networ...