检索结果-内蒙古大学图书馆

28th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD)

作者： Zhao, Yaru Huang, Yakun Cheng, Bo Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China

ISBN: (纸本)9789819722655;9789819722662

Despite the success of prompt learning-based models in text generation tasks, they still suffer from the introduction of external commonsense knowledge, especially from biased knowledge introduction. In this work, we propose KiProL, a knowledge-injected prompt learning framework to improve language generation and training efficiency. KiProL tackles ineffective learning and utilization of knowledge, reduces the biased knowledge introduction, as well as high training expenses. Then, we inject the recommended knowledge into the prompt learning encoder to optimize guiding prefixes without modifying the pre-trained model's parameters, resulting in reduced computational expenses and shorter training duration. Our experiments on two publicly available datasets (i.e., Explanation Generation and Story Ending Generation) show that KiProL outperforms baseline models. It improves fluency by an average of 2%, while diversity increases by 3.4% when compared with advanced prompt learning-based methods. Additionally, KiProL is 45% faster than the state-of-the-art knowledgeable, prompt learning method in training efficiency.

关键词： Language generation Prompt learning Knowledge injection Unbiased knowledge triples

来源：评论

学校读者我要写书评

暂无评论

Non-line-of-Sight Optical Positioning Under Full-Duplex Ultraviolet Communication Scenarios

Non-line-of-Sight Optical Positioning Under Full-Duplex Ultr...

引用

IEEE/CIC International Conference on Communications in China (ICCC)

作者： Liu, Gang Wang, Siming Yuan, Renzhi Peng, Mugen Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China

ISBN: (纸本)9798350378412

Determining the position of non-line-of-sight (NLOS) targets using traditional optical positioning methods remains a challenge. Recently, a NLOS optical positioning method using ultraviolet (UV) signals was proposed to enhance the single-scattering links of UV communications. Practical UV communication suffers from low achievable information rate due to the wide pulse spreading effect caused by strong scattering. Full-duplex techniques can be employed to improve the achievable information rate. However, full-duplex techniques introduce severe self-interference (SI) to transmitters of UV communications, which will inevitably degrade the performance of NLOS optical positioning. In this work, we propose a NLOS optical positioning method for full-duplex ultraviolet communications. To mitigate the influence of SI, we first approximate the path loss of the SI as a Fourier series (FS), then subtract the SI from the receiving power for calculating the location and pointing angles of the transmitter. Numerical results indicate that the proposed FS can will approximate the SI calculated from Monte-Carlo simulation models. Besides, we found that the proposed NLOS UV positioning method for full-duplex UV communications can achieve a positioning error of less than 3 meters and an azimuth error of less than 1.2 degrees within typical transceiver geometries up to 100 meters, significantly improving performance compared to positioning under scenarios with unresolved SI.

关键词： Full-duplex non-line-of-sight self-interference mitigation ultraviolet positioning

来源：评论

学校读者我要写书评

暂无评论

Delay and Energy Consumption Oriented UAV Inspection Business Collaboration Computing Mechanism in Edge Computing Based Electric Power IoT

引用

Chinese Journal of Electronics 2023年第1期32卷 13-25页

作者： SHAO Sujie LI Yi GUO Shaoyong WANG Chenhui CHEN Xingyu QIU Xuesong State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Blockchain Research Department China Electronics Standardization Institute

With the development of Internet of things(IoT) technology and smart grid infrastructure,edge computing has become an effective solution to meet the delay requirements of the electric power IoT. Due to the limitation of battery capacity and data transmission mode of IoT terminals, the business collaboration computing must consider the energy consumption of the terminals. Since delay and energy consumption are the optimization goals of two co-directional changes, it is difficult to find a business collaboration computing mechanism that simultaneously minimizes delay and energy *** paper takes the unmanned aerial vehicle(UAV) inspection business scenario in the electric power IoT based on edge computing as the representative, and proposes a two-stage business collaboration computing mechanism including resources allocation and task allocation to optimize the business delay and energy consumption of UAV by decoupling the complex correlation between resource allocation and task allocation. A steepest descent resource allocation algorithm is proposed. On the basis of resource allocation, an improved multiobjective evolutionary algorithm based on decomposition by dynamically adjusting the size of neighborhood and the cross distribution index is proposed as a task allocation algorithm to minimize energy consumption and business delay. Simulation results show that our algorithms can respectively reduce the business delay and energy consumption by more than6.4% and 9.5% compared with other algorithms.

关键词： Energy consumption Heuristic algorithms Collaboration Autonomous aerial vehicles Delays Power systems Resource management

来源：评论

学校读者我要写书评

暂无评论

Dynamic Resource Allocation for Remote IoT Data Collection in SAGIN

引用

IEEE INTERNET OF THINGS JOURNAL 2024年第11期11卷 20575-20589页

作者： Jia, Huaiqi Wang, Ying Wu, Wen Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China Peng Cheng Lab Frontier Res Ctr Shenzhen 518000 Peoples R China

In this article, we investigate a dynamic resource allocation problem for remote Internet of Things (IoT) data collection in space-air-ground integrated networks (SAGINs), in which the aerial platforms are deployed to bridge the communications between IoT nodes and satellites. To obtain an efficient resource allocation strategy that accommodates the stochastic data arrivals of IoT nodes and the dynamic network topology due to the high mobility of nongeostationary orbit (NGSO) satellites, we first formulate a resource allocation problem with queue stability constraints. Our objective is to maximize the long-term network utility, ensuring a balance between throughput and fairness among the IoT nodes. The formulated long-term problem is challenging to solve due to the unknown future network states and the coupling between continuous and integer variables. Therefore, we adopt the Lyapunov optimization theory to transform the problem into a deterministic problem in each time slot. Moreover, an online resource allocation algorithm is proposed to dynamically determine data admission, subchannel assignment, and power control in each time slot based on the current network status and data backlog. In addition, theoretical analysis indicates that there is an [O(1/V), O(V)] tradeoff between network utility and data backlog with control parameter V. Numerical results demonstrate that the proposed algorithm can greatly enhance the system throughput and reduce data queue backlog as well as preserve queue stability as compared with the benchmarks.

关键词： Internet of Things (IoT) Lyapunov optimization satellite communications space-air-ground integrated networks (SAGINs)

来源：评论

学校读者我要写书评

暂无评论

Age of Model: A Metric for Keeping AI Models Up-to-Date in Edge Intelligence-Enabled IoV

引用

IEEE TRANSACTIONS ON VEHICULAR technology 2024年第9期73卷 13047-13059页

作者： Chi, Xiaoyu Qin, Xiaoqi Xu, Xiaodong Han, Shujun Jin, Liang Zhang, Ping Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100150 Peoples R China Peng Cheng Lab Dept Broadband Commun Shenzhen 518066 Peoples R China

In the 6th generation (6G) mobile networks, intelligence is expected to be present in every part of the Internet of Vehicle (IoV) system to support smart transportation. In the context of constantly changing task requirements caused by dynamic environments of IoV, lifelong model update is essential to maintain high-quality inference performance. Given the mismatch between the resource demanding update process and limited system resource, it is critical to strike a balance between update quality and the update frequency. In this paper, we propose a performance metric called Age of Model (AoM) to characterize the inference performance degradation of AI models under dynamic task requirements. We derive the closed-form expressions of AoM at both edge tier and cloud tier, where various queues are exploited to abstract the model update process with single server or multiple servers. The obtained analytical framework describes the dependency among update request frequency, computing rate and training data diversity, which can serve as performance metric for system optimization. The accuracy and feasibility of our analysis are verified with extensive simulation results.

关键词： Measurement Data models Task analysis Servers Computational modeling Artificial intelligence Training Edge intelligence age of model AI model update multi-tier computing networks internet of vehicle

来源：评论

学校读者我要写书评

暂无评论

Distributed Intelligence Analysis Architecture for 6G Core Network 18th

Distributed Intelligence Analysis Architecture for 6G Core N...

引用

18th International Conference on Bio-Inspired Computing: Theories and Applications (BIC-TA)

作者： Sun, Wen Sun, QiBo Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China

ISBN: (纸本)9789819722747;9789819722754

To achieve automation and intelligence in 5G networks, the 3rd Generation Partnership Project (3GPP) introduced the Network Data Analysis Function (NWDAF) as a novel network function. However, in the traditional 5G core network architecture, NWDAF relies on fixed configurations for data collection, lacking support for user customization and flexibility. Additionally, the current deployment of NWDAF is predominantly centralized, failing to provide real-time and reliable analysis services for the massive data in future 6G systems. Moreover, it is incapable of ensuring user privacy, making it incompatible with emerging scenarios like federated learning in 6G. Therefore, this paper proposes a user-customizable data collection approach and introduces a distributed NWDAF deployment based on the Raft algorithm, where the master node assigns data collection, analysis, and inference tasks to multiple worker NWDAFs. Our work and experimental results demonstrate that the proposed architecture effectively addresses these challenges and further achieves closed-loop network automation in 6G systems.

关键词： Core network Network data analysis Raft algorithm

来源：评论

学校读者我要写书评

暂无评论

Network state Sensing Assisted Resource Allocation for Grant Free Multiple Access Systems with Users of Heterogeneous Delay Tolerance 59

Network State Sensing Assisted Resource Allocation for Grant...

引用

59th Annual IEEE International Conference on Communications (IEEE ICC)

作者： Li, Liujie Liu, Chenxi Cao, Bin Peng, Mugen Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China

ISBN: (纸本)9781728190549

In the design of grant-free multiple access (GFMA) mechanisms, the number of active users (UEs) with heterogeneous delay tolerance is commonly assumed to be known, which is less practical in real-world implementations. In this paper, we propose a network state sensing assisted resource allocation algorithm for the GFMA systems, in which both the delay-sensitive unmanned aerial vehicles (UAVs) and the delay-tolerant terrestrial UEs simultaneously access to the base stations (BSs). Taking the limited radio resources and the practical assumption that the number of active UEs is not available at the BSs into account, we first design a Bayesian-based estimation algorithm that can determine the number of delay-sensitive UAVs and delay-tolerant terrestrial UEs in real-time. Based on the estimated result, we further develop a dynamic algorithm to judiciously allocate the limited radio resources according to the weights and number between the UAVs and the terrestrial UEs. Moreover, we dynamically adjust the access class barring factors to ensure the access delay requirements of the UEs. Through numerical results, we show how our proposed algorithm can achieve almost the same performance as that of the ideal algorithm assuming perfect information on the number of active UEs is available.

关键词： Unmanned aerial vehicles (UAV)

来源：评论

学校读者我要写书评

暂无评论

Associative Tasks Computing Offloading Scheme in Internet of Medical Things with Deep Reinforcement Learning

引用

China Communications 2024年第4期21卷 38-52页

作者： Jiang Fan Qin Junwei Liu Lei Tian Hui School of Communications and Information Engineering Xi’an University of Posts and TelecommunicationsXi’an 710121China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing 100876China

The Internet of Medical Things(Io MT) is regarded as a critical technology for intelligent healthcare in the foreseeable 6G era. Nevertheless, due to the limited computing power capability of edge devices and task-related coupling relationships, Io MT faces unprecedented challenges. Considering the associative connections among tasks, this paper proposes a computing offloading policy for multiple-user devices(UDs) considering device-to-device(D2D) communication and a multi-access edge computing(MEC)technique under the scenario of Io MT. Specifically,to minimize the total delay and energy consumption concerning the requirement of Io MT, we first analyze and model the detailed local execution, MEC execution, D2D execution, and associated tasks offloading exchange model. Consequently, the associated tasks’ offloading scheme of multi-UDs is formulated as a mixed-integer nonconvex optimization problem. Considering the advantages of deep reinforcement learning(DRL) in processing tasks related to coupling relationships, a Double DQN based associative tasks computing offloading(DDATO) algorithm is then proposed to obtain the optimal solution, which can make the best offloading decision under the condition that tasks of UDs are associative. Furthermore, to reduce the complexity of the DDATO algorithm, the cacheaided procedure is intentionally introduced before the data training process. This avoids redundant offloading and computing procedures concerning tasks that previously have already been cached by other UDs. In addition, we use a dynamic ε-greedy strategy in the action selection section of the algorithm, thus preventing the algorithm from falling into a locally optimal solution. Simulation results demonstrate that compared with other existing methods for associative task models concerning different structures in the Io MT network, the proposed algorithm can lower the total cost more effectively and efficiently while also providing a tradeoff between delay and energy co

关键词： associative tasks cache-aided procedure double deep Q-network Internet of Medical Things(IoMT) multi-access edge computing(MEC)

来源：评论

学校读者我要写书评

暂无评论

Collaborative Multiple Access And Energy-Efficient Resource Allocation in Distributed Maritime Wireless Networks

引用

China Communications 2022年第4期19卷 137-153页

作者： Xueyan Cao Hongming Zhang Mugen Peng State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing 100876China

With the rapid increasing of maritime activities, maritime wireless networks(MWNs) with high reliability, high energy efficiency, and low delay are required. However, the centralized networking with fixed resource scheduling is not suitable for MWNs due to the special environment. In this paper,we introduce the collaborative relay communication in distributed MWNs to improve the link reliability, and propose an orthogonal time-frequency resource block reservation based multiple access(RRMA) scheme for both one-hop direct link and two-hop collaborative relay link to reduce the interference. To further improve the network performance, we formulate an energy efficiency(EE) maximization resource allocation problem and solve it by an iterative algorithm based on the Dinkelbach method. Finally, numerical results are provided to investigate the proposed RRMA scheme and resource allocation algorithm, showing that the low outage probability and transmission delay can be attained by the proposed RRMA scheme. Moreover,the proposed resource allocation algorithm is capable of achieving high EE in distributed MWNs.

关键词： maritime wireless networks distributed multiple access resource allocation energy efficiency maximization

来源：评论

学校读者我要写书评

暂无评论

SEMDA: COMMUNICATION-EFFICIENT DATA AGGREGATION THROUGH DISTRIBUTED SEMANTIC TRANSMISSION 49

SEMDA: COMMUNICATION-EFFICIENT DATA AGGREGATION THROUGH DIST...

引用

49th IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Zhao, Yaru Huang, Yakun Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China

ISBN: (纸本)9798350344868;9798350344851

This paper introduces SemDA, a communication-efficient data aggregation method that uses distributed semantic communication for improved transmission and analysis. SemDA utilizes an end-to-end trainable network structure that reduces data transmission volume and deepens semantic feature aggregation. key advances include an attention-based aggregation method for holistic semantic feature integration and a dual-attention decoding network that emphasizes viewpoint and content dimensions. Performance evaluations on CIFAR-10 and ImageNet datasets show that SemDA offers significant improvements in accuracy and system overhead compared to traditional and distributed semantic communication methods. Notable contributions include the proposal of a novel decoding structure, the introduction of a dual-attention decoding mechanism, and extensive evaluations against benchmark methods.

关键词： Data aggregation distributed semantic transmission dual attention communication-efficient

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：