检索结果-内蒙古大学图书馆

2023 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2023

作者： Luo, Yang Lin, Rongheng State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China

ISBN: (纸本)9798350337020

Long-tailed data distribution is a common phenomenon observed in nature. However, this distribution poses challenges for neural network training, as it performs well on samples with head categories, but not so well on tail categories - precisely what requires attention in certain scenes. Although existing methods involve re-sampling data or designing loss functions, they often fail to adapt well to changing data distributions. In this article, we propose a long-tailed detection approach based on multi-expert aggregation (MEA). This method employs multi-expert networks to learn about head, medium, and tail categories from a fixed long-tailed distribution. Different data are modelled using different loss designs, and the results of multi-expert networks are aggregated to achieve good accuracy for both head and tail categories. The effectiveness of this method has been verified on LVIS. © 2023 IEEE.

关键词： deep learning long-tailed recognition multi-expert learning object detection

来源：评论

学校读者我要写书评

暂无评论

Least Co-Residence: A Novel Virtual Migration Algorithm for Mitigating Side-Channel Attacks

Least Co-Residence: A Novel Virtual Migration Algorithm for ...

引用

2023 International Conference on networking and Network Applications, NaNA 2023

作者： Dong, Wei Zhou, Zan Yang, Shujie Lian, Yibo Li, Hongjing Xu, Changqiao Beijing University of Posts and Telecommunications State Key Laboratory of Networking and Switching Technology Beijing China

ISBN: (纸本)9798350327380

The advent of cloud platforms has revolutionized the landscape of modern business and user computing by providing a scalable and cost-effective alternative to the traditional model of purchasing and maintaining equipment. However, the coresidence feature, which is inherent to virtual machines (VMs) in cloud platforms, creates vulnerabilities that attackers can exploit to launch cache-based side-channel attacks, thereby compromising the security of user information. To address this problem, this paper proposes a hot migration algorithm based on a multiarmed bandit (MAB) model, which leverages physical machine (PM) migration between PMs to mitigate side-channel attacks in the cloud. Our proposed algorithm achieves the optimal timeaverage reward and ensures the security of the cloud platform, thereby outperforming other algorithms. The findings of this study have significant implications for the development of secure cloud computing systems, with the potential to enhance user privacy and safeguard against cyber threats. © 2023 IEEE.

关键词： Cost effectiveness

来源：评论

学校读者我要写书评

暂无评论

Logical-Cluster-Based Personalized Federated Multi-Task Learning for Internet of Vehicles

Logical-Cluster-Based Personalized Federated Multi-Task Lear...

引用

2023 IEEE International Symposium on Broadband Multimedia Systems and Broadcasting, BMSB 2023

作者： Zhang, Biao Xu, Siya Qiu, Xusong Tian, Jingyue Beijing University of Posts and Telecommunications State Key Laboratory of Networking and Switching Technology Beijing China

ISBN: (纸本)9798350321524

Federated learning (FL) is an emerging distributed machine learning paradigm that emphasizes user privacy. The majority of current federated learning systems are oriented for single task while participants in the Internet of Vehicles (IoV) demand to train different models for multiple intelligent services simultaneously. As one of transfer learning methods, multi-task learning (MTL) has the potential to integrate with federated learning to realize personalized local training. However, the existing federated multi-task learning (FMTL) algorithms are faced with the problems of high implementation complexity and communication overhead. To solve the above issues, we propose a logical-cluster-based personalized federated multi-task learning framework named pFMTL. In the framework, the multi-task model is decomposed into a basic module for extracting features and K task-specific modules for outputting inferences. We leverage logical clusters and multi-task learning to enhance the personalization and generalization capability of task models, respectively. To improve the communication efficiency further, we also design a module-wise task scheduling strategy, which supports both user module scheduling and cluster aggregation scheduling to ensure the convergence of multi-task model with less communication overhead. Finally, the simulation results imply that pFMTL can increase task accuracy and reduce communication latency compared with other benchmarks. © 2023 IEEE.

关键词： Vehicles

来源：评论

学校读者我要写书评

暂无评论

Service Migration in Mobile Edge Computing Based on 5G Network User Plane

Service Migration in Mobile Edge Computing Based on 5G Net...

引用

International Conference on Big Data Intelligence and Computing, DataCom 2022

作者： Hui, Liu QiBo, Sun State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China

ISBN: (纸本)9789819922321

The development of mobile edge computing under 5G network provides basic technical support for low latency and high reliability and also poses new challenges to service migration. 5G provides an uplink classifier (UL CL) function in data transmission, which makes it possible to build a new data transmission path. This paper proposes a new service migration decision model based on the 5G network user plane. Based on the new service migration model, this paper uses Markov decision process to construct service migration decision problem, and proposes an online service migration algorithm based on reinforcement learning to learn the optimal migration strategy. Finally, we compare the proposed algorithm with other benchmark algorithms. The results show that the online algorithm is close to the global performance, superior to the existing benchmark algorithm in terms of problem formulation proposed in our paper. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： 5G mobile communication systems

来源：评论

学校读者我要写书评

暂无评论

Stable 5G Time Domain Resource Configuration for Synchronous Timing Services via Lyapunov Aided DRL 36

Stable 5G Time Domain Resource Configuration for Synchronous...

引用

36th IEEE/IFIP Network Operations and Management Symposium, NOMS 2023

作者： Zhou, Yanbo Feng, Lei Xie, Kunyi Zhou, Fanqin Li, Wenjing Yu, Peng Beijing University of Posts and Telecommunications The State Key Laboratory of Networking and Switching Technology Beijing China

ISBN: (纸本)9781665477161

The high-precision clock synchronization is pursued with the consideration of the balance for time-domain resource utilization, when 5G technologies are expected to carry the timing services. Firstly, a clock synchronization model is established in the case that the system exists the observed value loss. The probability distribution of the loss of the observed value is evaluated by a newly proposed delay deterministic confidence method. Then, the error boundness of the clock synchronization is investigated by the Kalman filter algorithm, and the optimization problem for 5G time-domain resource configuration is formulated with guaranteeing the precision of clock synchronization;Finally, the proposed optimization problem is solved by dueling double deep Q-learning-based Lyapunov optimization. The experimental results verify the effectiveness and superiority of the proposed method in terms of the joint optimization of clock synchronization error covariance and throughput. © 2023 IEEE.

关键词： Optimization

来源：评论

学校读者我要写书评

暂无评论

NeRFCom: Feature Transform Coding Meets Neural Radiance Field for Free-View 3-D Scene Semantic Transmission

引用

IEEE COMMUNICATIONS LETTERS 2025年第5期29卷 898-902页

作者： Yue, Weijie Si, Zhongwei Wu, Bolin Wang, Sixian Qin, Xiaoqi Niu, Kai Dai, Jincheng Zhang, Ping Beijing Univ Posts & Telecommun Key Lab Universal Wireless Commun Minist Educ Beijing 100876 Peoples R China Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China

We introduce NeRFCom, a novel communication system designed for end-to-end 3D scene transmission. Compared to traditional systems relying on handcrafted NeRF semantic feature decomposition for compression and well-adaptive channel coding for transmission error correction, our NeRFCom employs a nonlinear transform and learned probabilistic models, enabling flexible variable-rate joint source-channel coding and efficient bandwidth allocation aligned with the NeRF semantic feature's different contribution to the 3D scene synthesis fidelity. Experimental results demonstrate that NeRFCom achieves free-view 3D scene efficient transmission while maintaining robustness under adverse channel conditions.

关键词： Three-dimensional displays Neural radiance field Symbols Feature extraction Decoding Entropy Optimization Image reconstruction Encoding Transforms Neural radiance field (NeRF) 3D scene transmission semantic features nonlinear transform coding

来源：评论

学校读者我要写书评

暂无评论

SLIT: Achieving Fast Bandwidth Isolation Across Virtual Machines

引用

IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT 2023年第3期20卷 3004-3019页

作者： Huang, Chengyuan Zhang, Jiao Huang, Tao Purple Mt Labs Nanjing 211111 Peoples R China Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China

Network performance guarantee in the cloud is one of the hottest research topics recently. However, current approaches either lack scalability or fail to achieve high performance. To avoid the hard tradeoff between the scalability and performance, we adopt a hybrid approach to get the best of both worlds. In this paper, we follow the stateless design principle and propose SLIT, which provides fast VM-level network performance isolation and maintains core-stateless characteristics. Specifically, at the end host, we develop a novel network-adaptive labeling mechanism and it computes the correct scheduling priority for each hop. At the switch, packets are scheduled based on their labels and we develop a novel straggler detection algorithm to find the new incoming flow. Our evaluation results show that SLIT can approximate the optimal performance of Weighted Fair Queuing (WFQ) effectively.

关键词： Cloud computing guaranteed network performance core-stateless bandwidth isolation

来源：评论

学校读者我要写书评

暂无评论

Blockchain-based Pricing Mechanism Research on IoT Data Transactions 29

Blockchain-based Pricing Mechanism Research on IoT Data Tran...

引用

29th IEEE International Conference on Parallel and Distributed Systems, ICPADS 2023

作者： Ding, Shiqi Guo, Shaoyong Liu, Chang Wang, Zhili Beijing University of Posts and Telecommunications The State Key Laboratory of Networking and Switching Technology Beijing China

ISBN: (纸本)9798350330717

At the current stage, data trading is still in the initial stage, there is no mature pricing mechanism, and it is difficult to realize the effective use of data *** order to incentivize the better circulation of data and give full play to the value of data, it is necessary to build a pricing mechanism for data transactions between data owners and data requesters. To address the above issues, this paper proposes a blockchain-based data transaction pricing mechanism. First, for the privacy and security of data, we adopt a blockchain-based architecture to guarantee the security of the transaction data and at the same time satisfy the arbitration of data transactions. Second, we formulate a pricing scheme based on data quality to provide a basis for further game dynamic pricing;then, in this paper, in order to incentivize both parties of the transaction to actively participate in the transaction and maximize the utility of both parties as much as possible, we construct a pricing model based on the Stackelberg game;lastly, we design an improved Double Deep Q Network (DDQN) reinforcement learning algorithm to optimize the strategy of pricing in data transactions. Experiments show that the pricing mechanism proposed in this paper improves the utility of both parties in the data pricing process, and improves the total benefit by 10% compared to the DQN algorithm. © 2023 IEEE.

关键词： Blockchain Data trading Reinforcement learning Stackelberg game

来源：评论

学校读者我要写书评

暂无评论

DT-PBFT: A Double-Layer Group Consensus Algorithm of Credibility for IoT Blockchain 2

DT-PBFT: A Double-Layer Group Consensus Algorithm of Credibi...

引用

2nd International Conference on Big Data, Information and Computer Network, BDICN 2023

作者： Chen, Yaxin Jia, Yongpu Beijing University of Posts and Telecommunications State Key Laboratory of Networking and Switching Technology Beijing China

ISBN: (纸本)9798350321364

Blockchain technology is becoming more and more popular, but performance problems have always troubled it, especially the scale of distributed networks is limited. Like Practical Byzantine consensus algorithm (PBFT), the main bottlenecks are concentrated in the scope of consensus too small, and the communication is too high. In order to implement PBFT in large-scale systems such as large-scale IoT ecosystems and blockchains, this paper proposes a reputation-based two-layer improved PBFT consensus computing model DT-PBFT. In order to better combine with the large -scale Internet of Things ecosystem, we give each node regional location, communication capacity, computing power and other related attributes to each involvement, so that Client nodes can input node attributes as parameters into the reputation calculation algorithm, to select the master node of each layer in the double-layer consensus model. Determining the master node through reputation improves the credibility of the consensus algorithm. At the same time, the DTPBFT cancels the Pre-Prepare stage, reducing communication complexity. Finally, this paper transforms the existing PBFT algorithm on the Fabric platform to implement the DT-PBFT consensus algorithm and compares its performance with traditional PBFT, proving that DT-PBFT has the best balance and excellent comprehensive performance in large-scale IoT application scenarios. © 2023 IEEE.

关键词： Blockchain

来源：评论

学校读者我要写书评

暂无评论

Dense Contrastive-Based Federated Learning for Dense Prediction Tasks on Medical Images

引用

IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS 2024年第4期28卷 2055-2066页

作者： Yang, Yuning Liu, Xiaohong Gao, Tianrun Xu, Xiaodong Zhang, Ping Wang, Guangyu Beijing Univ Posts & Telecommun State Key Lab Networking & Switching Technol Beijing 100876 Peoples R China UCL UCL Canc Inst London WC1E 6DD England

Deep learning (DL) models have achieved remarkable success in various domains. But training an accurate DL model requires large amounts of data, which can be challenging to obtain in medical settings due to privacy concerns. Recently, federated learning (FL) has emerged as a promising solution that shares local models instead of raw data. However, FL in medical settings faces challenges of client drift due to the data heterogeneity across dispersed institutions. Although there exist studies to address this challenge, they mainly focus on the classification tasks that learn global representation of an entire image. Few have been studied on the dense prediction tasks, such as object detection. In this study, we propose dense contrastive-based federated learning (DCFL) tailored for dense prediction tasks in FL settings. DCFL introduces dense contrastive learning to FL, which aligns the local optimization objectives towards the global objective by maximizing the agreement of representations between the global and local models. Moreover, to improve the performance of dense target prediction at each level, DCFL applies multi-scale contrastive representation by utilizing multi-scale representations with dense features in contrastive learning. We evaluated DCFL on a set of realistic datasets for pulmonary nodule detection. DCFL demonstrates an overall performance improvement compared with the other federated learning methods in heterogeneous settings-improving the mean average precision by 4.13% and testing recall by 6.07% in highly heterogeneous settings.

关键词： Task analysis Predictive models Training Federated learning Self-supervised learning Prediction algorithms Servers Client drift dense prediction object detection federated learning contrastive learning

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：