检索结果-内蒙古大学图书馆

23rd IEEE International Conference on High Performance Computing and Communications, 7th IEEE International Conference on Data Science and Systems, 19th IEEE International Conference on Smart City and 7th IEEE International Conference on Dependability in Sensor, Cloud and Big Data Systems and Applications, HPCC-DSS-SmartCity-DependSys 2021

作者： Song, Yipei Lin, Rongheng Zou, Hua Beijing University of Posts and Telecommunications State Key Lab of Networking and Switching Technology Beijing100876 China

ISBN: (纸本)9781665494571

Information extraction aims to provide required high-quality RDF triples for knowledge graph which is the basis of knowledge construction. The current two mainstream methods include a joint model concentrates on using deep learning to extract entities and relationships and a pipeline model based on sequential extraction. The joint method will be limited by the number of parameters and the types of parameters, this will lead to the loss of calculation results. In the pipeline model, the result of entity extraction needs to be substituted into the relationship classification, and this usually causes the interference of redundant entities to the relationship classification. For the joint training model, in the training process, entities and relationships are extracted at the same time. The results of extractions are continuously transmitted to each other during training and influence each other. If a certain parameter limit is deleted during the extraction of one type of result, the extraction result of the other side will also be deleted of the same type, which will eventually cause some results to be lost. For the pipeline model, if entities are first extracted, then relations are classified. In the process of entity extraction, invalid entity pairs with no matching relationship will be generated. This causes the convolution training speed to slow down. In this paper, in order to solve such problems in entity and relationship extraction, we propose an entity relationship extraction model includes Dilated Gate unit convolutional layers with Weight Pointer Networks, named this structure as DGWPN. This is helpful to obtain more receptive fields and multiple triples models. The advantages of our model are demonstrated through the comparison of multiple groups of experiments. The experimental results show that our method outperforms other current different baseline models on the SKE and CHIP data sets, about 5.3% higher than the current mainstream Bi-LSTM with CRF model o

关键词： Training Knowledge engineering Deep learning Pipelines Interference Logic gates Information retrieval

来源：评论

学校读者我要写书评

暂无评论

Resource allocation and cost optimization in relay-assisted mobile edge computing

引用

COMPUTING 2025年第5期107卷 1-25页

作者： Zhan, Huifang Wu, Guilu Li, Zhengquan Nie, Gaofeng Jiangnan Univ Sch Internet Things Engn Wuxi Peoples R China Beijing Univ Posts & Telecommun State key Lab Networking & Switching Technol Beijing Peoples R China Hainan Univ Sch Informat & Commun Engn Hainan Peoples R China

Motivated by the limited computing and storage resources of smart devices (SDs), existing resource management schemes under the traditional mobile edge computing (MEC) system couldn't fully utilize the network resources. This article proposes a novel relay-assisted MEC architecture to provision SDs, where an idle relay device assumes the dual role of providing fee-based communication and computing services. At the same time, tasks generated by applications exhibit three distinct processing modes: local execution, relay offloading, and edge computing. To minimize the execution cost associated with communication and computing services from the user's perspective, joint computation offloading, relay selection strategy, computation resource allocation, and spectrum resource allocation (JORA) problem is formulated, which is a non-convex problem. Subsequently, a two-stage iterative algorithm is designed to solve the JORA problem using block coordinate descent, reconstruction linearization technology, and convex optimization methods. Simulation results show the effectiveness of the proposed JORA algorithm, which can significantly save the execution cost for edge users.

关键词： Mobile edge computing Relay-assisted computing offloading Execution cost Resource allocation Block coordinate descent

来源：评论

学校读者我要写书评

暂无评论

QoS-Aware Service Composition Method Based on Double Channel Action Network Reinforcement Learning 14th

QoS-Aware Service Composition Method Based on Double Channel...

引用

14th International Conference on Computer Engineering and Networks, CENet 2024

作者： Guo, Yanpeng Song, Chunxiao Huang, Ruoyan Yang, Yang State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China The 54th Research Institute of CETC Hebei Shijiazhuang China

ISBN: (纸本)9789819649624

Currently, the problem of Quality-of-service in service composition has emerged as a critical research topic in the fields of cloud computing and service computing. Aggregating multiple existing services based on QoS awareness to meet complex functional and non-functional requirements, which simple services cannot satisfy, and choosing services for aggregation in a rapidly changing service environment to achieve the highest possible service quality, poses an NP-hard challenge. Traditional heuristic algorithms, while efficient to an extent, often show limitations in adapting to this rapidly changing environment as they typically rely on static rules or preset conditions, lacking the ability to learn from and adapt to environmental changes. Methods based on reinforcement learning are the most commonly used approach in dynamic and complex environments. However, as the scale of composition increases, they face challenges such as constantly changing service characteristics, user demands, and a vast state space. Therefore, this paper designs a service composition framework based on graph convolution and Deep Reinforcement Learning (DRL), which extracts environmental states from three dimensions: abstract services, specific services, and QoS features, using graph convolution, makes decisions using DRL, and reduces the state space with a double channel action network. Simulations have demonstrated that our method surpasses existing methods in terms of service composition quality, scalability, and stability. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

HAP-assisted multi-aerial base station deployment for capacity enhancement via federated deep reinforcement learning

引用

JOURNAL OF CLOUD COMPUTING-ADVANCES SYSTEMS AND APPLICATIONS 2023年第1期12卷 136-136页

作者： Liu, Lei He, Haoran Qi, Fei Zhao, Yikun Xie, Weiliang Zhou, Fanqin Feng, Lei China Telecom Corp Ltd Beijing Res Inst Beijing 102209 Peoples R China Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China

Aerial base stations (AeBSs), as crucial components of air-ground integrated networks, are widely employed in cloud computing, disaster relief, and various applications. How to quickly and efficiently deploy multi-AeBSs for higher capacity gain has become a key research issue. In this paper, we address the 3D deployment optimization problem of multi-AeBSs with the objective of maximizing system capacity. To overcome communication overhead and privacy challenges in multi-agent deep reinforcement learning (MADRL), we propose a federated deep deterministic policy gradient (Fed-DDPG) algorithm for the multi-AeBS deployment decision. Specifically, a high-altitude platform (HAP)-assisted multi-AeBS deployment architecture is designed, in which low-altitude AeBS act as the local nodes to train its own deployment decision model, while the HAP acts as the global node to aggregate the weights of local models. In this architecture, AeBSs do not exchange raw data, addressing data privacy concerns and reducing communication overhead. Simulation results show that the proposed algorithm outperforms fully distributed MADRL algorithms and closely approximates the performance of multi-agent deep deterministic policy gradient (MADDPG), which requires global information during training, but with less training time.

关键词： Aerial base station (AeBS) Capacity enhancement Deep reinforcement learning (DRL) Federated reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Explain-Analyze-Generate: A Sequential Multi-Agent Collaboration Method for Complex Reasoning 31

Explain-Analyze-Generate: A Sequential Multi-Agent Collabora...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Gu, Wenyuan Han, Jiale Wang, Haowen Li, Xiang Cheng, Bo State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications China Hong Kong University of Science and Technology Hong Kong School of Computer Science and Technology Anhui University China

ISBN: (纸本)9798891761964

Exploring effective collaboration among multiple large language models (LLMs) represents an active research direction, with multiagent debate (MAD) emerging as a popular approach. MAD involves LLMs independently generating responses and refining their own responses by incorporating feedback from other agents in a debate manner. However, empirical experiments reveal the suboptimal performance of MAD in complex reasoning scenarios. We attribute this to the potential misleading caused by peer agents with limited individual capabilities. To address this, we propose a novel sequential collaboration framework named Explain-Analyze-Generate (EAG). By decomposing complex tasks into essential subtasks and employing a pipeline approach, EAG enable agents provide constructive assistance to peers, ultimately yielding higher performance. We conduct experiments on the comprehensive complex language reasoning benchmark: BIG-Bench-Hard (BBH). Our method achieves the highest performance on 19 out of 23 tasks, with an average improvement of 8% across all tasks, and incurs lower costs compared to MAD, demonstrating its effectiveness and efficiency. © 2025 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Network Intrusion Detection System Based on Domain Adaptation for Industrial Control System 5

Network Intrusion Detection System Based on Domain Adaptatio...

引用

5th International Conference on Artificial Intelligence Technologies and Applications, ICAITA 2023

作者： Zhang, Jiahui Yu, Le Zhang, Feng State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing100876 China China Mobile Communications Corporation Beijing100053 China

ISBN: (纸本)9781643684840

Industrial control systems play a crucial role in society, and ensuring the security of industrial control systems has become an important and challenging task. Intrusion detection is an important defense mechanism for protecting these systems. However, in existing industrial control intrusion detection techniques, especially those based on deep learning, a large number of training samples are usually required, and real-world samples often exhibit data imbalance, which can affect the training effectiveness. In this paper, a network intrusion detection model based on domain adaptation is proposed, which is a sample-discriminating adversarial domain adaptation model that can classify all samples of the target domain in a universal domain adaptation scenario. © 2024 The Authors.

关键词： Intrusion detection

来源：评论

学校读者我要写书评

暂无评论

Two-Step Amplitude Barycenter Calibrations for OTFS Signals: A High-Resolution Sensing Waveform Design

引用

IEEE TRANSACTIONS ON VEHICULAR technology 2024年第12期73卷 18923-18936页

作者： Liu, Xiqing Gong, Jialong Yang, Yiran Wang, Dong Peng, Mugen Beijing Univ Posts & Telecommun Sch Informat & Commun Engn State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China China Telecom Res Inst Res Ctr 6G Beijing 100035 Peoples R China

Sensing is considered a native capability of the sixth-generation (6G) mobile communication systems, inspiring a variety of research endeavors with regard to integrated sensing and communication (ISAC) waveforms. The orthogonal time frequency space (OTFS) is a promising modulation technique for high-mobility ISAC applications. OTFS accommodates the channel dynamics in the delay-Doppler (DD) domain, where the delay and Doppler shifts of each physical path will be captured by the DD grid and represented as integer and fractional indexes. However, the critical challenge associated with OTFS-based sensing is the insufficient resolution caused by the limited bandwidth and frame duration. To address this issue, we propose a two-step amplitude barycenter calibration (TS-ABC) method, which exploits the non-linear relationship among several discrete Fourier transform (DFT) samples and produces a true-valued estimate through two calibration steps. The derivation and simulation results demonstrate that the TS-ABC is a low-complexity method compared to the existing estimators, and it yields superior sensing resolution.

关键词： Symbols Doppler effect Accuracy Delays Calibration OFDM Interference Amplitude barycenter calibration ISAC resolution OTFS

来源：评论

学校读者我要写书评

暂无评论

Genuine Einstein-Podolsky-Rosen steering of generalized three-qubit states via unsharp measurements

引用

Chinese Physics B 2023年第4期32卷 215-225页

作者：陈玉玉郭奋卓魏士慧温巧燕 School of Science Beijing University of Posts and TelecommunicationsBeijing 100876China Henan Key Laboratory of Network Cryptography Technology Zhengzhou 450001China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing 100876China CETC Cyberspace Security Technology Co. Ltd.Beijing 100041China

We aim to explore all possible scenarios of(1→2)(where one wing is untrusted and the others two wings are trusted)and(2→1)(where two wings are untrusted,and one wing is trusted)genuine tripartite Einstein-Podolsky-Rosen(EPR)*** generalized Greenberger-Horne-Zeilinger(GHZ)state is shared between three spatially separated parties,Alice,Bob and *** both(1→2)and(2→1),we discuss the untrusted party and trusted party performing a sequence of unsharp measurements,*** each scenario,we deduce an upper bound on the number of sequential observers who can demonstrate genuine EPR steering through the quantum violation of tripartite steering *** results show that the maximum number of observers for the generalized GHZ states can be the same with that of the maximally GHZ state in a certain range of state ***,both the sharpness parameters range and the state parameters range in the scenario of(1→2)steering are larger than those in the scenario of(2→1)steering.

关键词： quantum information quantum steering generalized Greenberger-Horne-Zeilinger(GHZ)state

来源：评论

学校读者我要写书评

暂无评论

Energy Consumption Minimization for NOMA-Based Secure UAV-MEC Network

引用

China Communications 2025年第3期22卷 202-216页

作者： Zhang Hao Huang Yuzhen Zhang Zhi Lu Xingbo School of Information and Communication Engineering Beijing University of Posts and TelecommunicationsBeijing 100876China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing 100876China The Academy of Military Sciences of PLA Beijing 100091China

Applying non-orthogonal multiple access(NOMA)to the mobile edge computing(MEC)network supported by unmanned aerial vehicles(UAVs)can improve spectral efficiency and achieve massive user access on the basis of solving computing resource constraints and coverage ***,the UAV-enabled network has a serious risk of information leakage on account of the openness of wireless *** paper considers a UAV-MEC secure network based on NOMA technology,which aims to minimize the UAV energy *** achieve the purpose while meeting the security and users’latency requirements,we formulate an optimization problem that jointly optimizes the UAV trajectory and the allocation of network *** that the original problem is non-convex and multivariate coupled,we proposed an effective algorithm to decouple the nonconvex problem into independent user relation coefficients and subproblems based on successive convex approximation(SCA)and block coordinate descent(BCD).The simulation results showcase the performance of our optimization scheme across various parameter settings and confirm its superiority over other benchmarks with respect to energy consumption.

关键词： MEC NOMA resource optimization secure transmission trajectory optimization UAV

来源：评论

学校读者我要写书评

暂无评论

A Trusted Edge Resource Allocation Framework for Internet of Vehicles

引用

Computers, Materials & Continua 2023年第11期77卷 2629-2644页

作者： Yuxuan Zhong Siya Xu Boxian Liao Jizhao Lu Huiping Meng Zhili Wang Xingyu Chen Qinghan Li State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing100876China State Grid Henan Information&Telecommunication Company Zhengzhou450018China School of Information Studies Syracuse UniversitySyracuse13244USA

With the continuous progress of information technique,assisted driving technology has become an effective technique to avoid traffic *** to the complex road conditions and the threat of vehicle information being attacked and tampered with,it is difficult to ensure information *** paper uses blockchain to ensure the safety of driving information and introduces mobile edge computing technology to monitor vehicle information and road condition information in real time,calculate the appropriate speed,and plan a reasonable driving route for the *** solve these problems,this paper proposes a trusted edge resource allocation framework for assisted driving service,which includes two stages:the blockchain generation stage(the first stage)and assisted driving service stage(the second stage).Furthermore,in the first stage,a delay-and-throughput-oriented block generation model for the mobile terminal is *** the second stage,a balanced offloading algorithm for assisted driving service based on edge collaboration is proposed to solve the problems of unbalanced load of cluster mobile edge computing(MEC)servers and low resource utilization of the *** this paper optimizes the throughput of blockchain and delay of the transportation network through deep reinforcement learning(DRL)***,compared with joint computation and communication resources’allocation(JCCR)and resource allocation method based on binary offloading(RAB),our proposed scheme can optimize the delay by 7.4%and 26.7%,and support various application services of the vehicular networks more effectively.

关键词： Blockchain load balancing vehicular networks resource allocation

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：