检索结果-内蒙古大学图书馆

Application-Oriented Cloud Workload Prediction:A Survey and New Perspectives

Tsinghua Science and Technology 2025年第1期30卷 34-54页

作者： Binbin Feng Zhijun Ding Key Laboratory of Embedded System and Service Computing Ministry of EducationDepartment of Computer Science and TechnologyTongji UniversityShanghai 201804China Key Laboratory of Embedded System and Service Computing Ministry of EducationTongji UniversityDepartment of Computer Science and TechnologyTongji UniversityShanghai 201804China Shanghai Artificial Intelligence Laboratory Shanghai 200232China

Workload prediction is critical in enabling proactive resource management of cloud *** workload prediction is valuable for cloud users and providers as it can effectively guide many practices,such as performance assurance,cost reduction,and energy consumption ***,cloud workload prediction is highly challenging due to the complexity and dynamics of workloads,and various solutions have been proposed to enhance the prediction *** paper aims to provide an in-depth understanding and categorization of existing solutions through extensive literature *** existing surveys,for the first time,we comprehensively sort out and analyze the development landscape of workload prediction from a new perspective,i.e.,application-oriented rather than prediction methodologies per ***,we first introduce the basic features of workload prediction,and then analyze and categorize existing efforts based on two significant characteristics of cloud applications:variability and ***,we also investigate how workload prediction is applied to resource ***,open research opportunities in workload prediction are highlighted to foster further advancements.

关键词： cloud computing workload prediction resource management artificial intelligence for IT operations(AIOps)

来源：评论

学校读者我要写书评

暂无评论

3DGCformer: 3-Dimensional Graph Convolutional transformer for multi-step origin–destination matrix forecasting

引用

Applied Intelligence 2025年第6期55卷 1-18页

作者： Huang, Yiou Deng, Hao Zhao, Shengjie School of Computer Science and Technology Tongji University Shanghai China Key Laboratory of Embedded System and Service Computing Shanghai China Engineering Research Center of Key Software Technologies for Smart City Perception and Planning Ministry of Education Shanghai China

Forecasting Human mobility is of great significance in the simulation and control of infectious diseases like COVID-19. To get a clear picture of potential future outbreaks, it is necessary to forecast multi-step Origin–Destination (OD) matrices for a relatively long period in the future. However, multi-step Origin–Destination Matrix Forecasting (ODMF) is a non-trivial problem. First, previous ODMF models only forecast the OD matrix for the next time-step, and they cannot perform well on long-term multi-step forecasts due to error accumulation. Second, many ODMF methods capture spatial and temporal dependencies with separate modules, which is insufficient to model spatio-temporal correlations in the time-varying OD matrix sequence. To address the challenges in multi-step ODMF, we propose 3-Dimensional Graph Convolutional Transformer (3DGCformer). As an enhancement of the original 3DGCN, we propose a novel Origin–Destination Feature Propagation (ODFP) rule between 3DGCN layers and integrate 2 3DGCNs with different spatio-temporal graphs and corresponding feature propagation rules to model the formation of OD flows in a more comprehensive way. For multi-step forecasts, 3DGCformer uses Transformer to capture long-term global temporal dependency, and adapt its decoder using labeled tokens to avoid error accumulation and improve time efficiency. To avoid information loss as the number of regions increases, we propose a patch embedding approach to convert data from 3DGCNs to the Transformer module. We perform extensive experiments on 4 real-world human mobility datasets, and the results show that our proposed model outperforms the state-of-the-art methods. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2025.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

UAV-Mounted Intelligent Reflecting Surface in Maritime Wireless Powered Communication Network

引用

China Communications 2025年第3期22卷 254-269页

作者： Liu Yuan Han Fengxia Zhao Shengjie School of Information Science and Electrical Engineering Shandong Jiaotong UniversityJinan 250357China College of Electronic and Information Engineering Tongji UniversityShanghai 201804China School of Software Engineering Tongji UniversityShanghai 201804China Engineering Research Center of Key Software Technologies for Smart City Perception and Planning Ministry of EducationShanghai 200003China Key Laboratory of Embedded System and Service Computing Tongji University Shanghai 201804China

Intelligent reflecting surface(IRS)assisted with the wireless powered communication network(WPCN)can enhance the desired signal energy and carry out the power-sustaining problem in ocean monitoring *** this paper,we investigate a reliable communication structure where multiple buoys transmit data to a base station(BS)with the help of the unmanned aerial vehicle(UAV)-mounted IRS and harvest energy from the base station *** organically combine WPCN with maritime data collection scenario,a scheduling protocol that employs the time division multiple access(TDMA)is proposed to serve multiple buoys for uplink data ***,we compare the full-duplex(FD)and half-duplex(HD)mechanisms in the maritime data collection system to illustrate different performances under these two *** maximize the fair energy efficiency under the energy harvesting constraints,a joint optimization problem on user association,BS transmit power,UAV’s trajectory and IRS’s phase shift is *** solve the non-convex problem,the original problem is decoupled into several subproblems,and successive convex optimization and block coordinate descent(BCD)methods are employed obtain the near-optimal solutions *** results demonstrate that the UAV-mounted IRS can significantly improve energy efficiency in our considered system.

关键词： full-duplex IRS joint optimization TDMA UAV WPCN

来源：评论

学校读者我要写书评

暂无评论

Binary-Encoding-Based Quantized Kalman Filter: An Approximate MMSE Approach

引用

IEEE Transactions on Automatic Control 2025年第5期70卷 3181-3196页

作者： Liu, Qinyuan Nie, Yao Wang, Zidong Dong, Hongli Jiang, Changjun Tongji University School of Computer Science and Technology Shanghai201804 China Tongji University Key Laboratory of Embedded System and Service Computing Ministry of Education Shanghai201804 China Tongji University Shanghai Artificial Intelligence Laboratory Shanghai201804 China Brunel University London Department of Computer Science UxbridgeUB8 3PH United Kingdom Northeast Petroleum University Artificial Intelligence Energy Research Institute Daqing163318 China Northeast Petroleum University Heilongjiang Provincial Key Laboratory of Networking and Intelligent Control Daqing163318 China

In this article, the Kalman filter design problem is investigated for linear discrete-time systems under binary encoding schemes. Under such a scheme, the local information is quantized into a bit string by the remote sensor based on a probabilistic quantizer, and then the bit string is transmitted via memoryless binary symmetric channels (BSCs). Due to the communication link noises, the bit flipping occurs in a random manner, and thus, the transmission of the bit string would suffer from specific bit-error rates. With the received bits, a recursive binary-encoding-based quantized Kalman filter is established in the approximate minimum mean-square error (MMSE) sense, which relies on the Gaussian approximation of the conditional probability density function at each iteration. Furthermore, the proposed estimator is shown to be of a Kalman-like type through performance analysis, which exhibits computational complexity comparable to the conventional Kalman filter. Subsequently, a posterior Cramér-Rao lower bound is derived for the proposed binary-encoding-based quantized Kalman filter. The effectiveness of the proposed estimator is demonstrated through numerical results. © 1963-2012 IEEE.

关键词： Mean square error

来源：评论

学校读者我要写书评

暂无评论

A two-way dynamic adaptive pricing resource allocation model based on combinatorial double auctions in computational network

引用

Computer Communications 2025年 238卷

作者： Xu, Yanjun Tian, Chunqi Wang, Wei Bai, Lizhi Xia, Xuhui School of Computer Science and Technology Tongji University Shanghai China School of Data Science and Engineering East China Normal University Shanghai China Embedded System and Service Computing Key Laboratory of Ministry of Education China

Efficient resource allocation in computing networks is essential for managing fluctuating demands and optimizing system performance. Traditional auction and pricing models often fail to adapt to diverse demands and supply–demand fluctuations, resulting in inefficiencies. This paper proposes a bidirectional price-adaptive bundled resource auction model that considers not only the autonomous adjustment of sellers’ quotations in response to supply–demand fluctuations but also the impact of these fluctuations on buyers’ willingness to pay and bidding behavior. The model integrates Combinatorial Double Auction (CDA) mechanism and Genetic Algorithm (GA), constructing a bundled resource auction mechanism that accommodates diverse resource demands and an adaptive pricing strategy that dynamically responds to real-time supply–demand variations. This approach enhances resource allocation accuracy in dynamic and competitive computing network environments. Furthermore, a reserve price mechanism and a delay compensation strategy are introduced to ensure that the proposed mechanism satisfies individual rationality, budget balance, and incentive compatibility while maintaining computational efficiency. Simulation results demonstrate that, compared to traditional methods, the proposed model not only improves allocation efficiency and enhances resource utilization but also helps reduce operational costs. Specifically, resource allocation dispersion decreases by 4.26%, while service providers’ revenue increases by 7.47%. This study provides a scalable and adaptive solution for dynamic resource allocation in cloud and edge computing platforms. It contributes significantly to the development of resource management and flexible pricing strategies in markets characterized by diverse demands and fluctuating conditions. © 2025

关键词： Budget control

来源：评论

学校读者我要写书评

暂无评论

Personalized Medical Federated Learning Based on Mutual Knowledge Distillation in Object Heterogeneous Environment 1st

Personalized Medical Federated Learning Based on Mutual Kno...

引用

1st Conference on Blockchain and Web3.0 Technology Innovation and Application, BWTAC 2024

作者： Ni, Lina Song, Chenglin Zhao, Hanmo Tang, Yuncan Ma, Yunshen Zhang, Jinquan College of Computer Science and Engineering Shandong University of Science and Technology Qingdao266590 China Key Laboratory of the Ministry of Education for Embedded System and Service Computing Tongji University Shanghai201804 China

ISBN: (纸本)9789819794119

Federated Learning (FL), hailed as a potent approach in merging medical expertise, promises to elevate collaborative efforts among healthcare institutions while safeguarding the privacy and security of sensitive medical data, thereby energizing the trajectory of intelligent healthcare advancements. However, the diversity in training objectives across medical institutions can diminish their active participation in a unified federated training process. To address this issue of waning enthusiasm due to diverse training goals, in this paper, we propose Personalized Medical Federated Learning method based on Mutual Distillation Knowledge (pFedMKD). Specifically, we propose a similarity-based mutual distillation model selection algorithm that reduces interference among medical institutions with notably diverse data distributions. Moreover, we develop a server-side local model update method leveraging mutual knowledge distillation, enabling each local model to gain beneficial insights for personalization from analogous models. Additionally, we formulate a personalized model update method utilizing knowledge distillation from specifically labeled samples, thereby improving the performance of personalized models. Experimental results demonstrate that compared to the baseline algorithm, pFedMKD can more effectively address the target heterogeneity problem among different medical institutions, enabling each institution to achieve higher personalized performance in federated learning tasks. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Adaptive Federated Learning Based on Device Performance in a Heterogeneous Environment of Medical computing Devices 1st

Adaptive Federated Learning Based on Device Performance in...

引用

1st Conference on Blockchain and Web3.0 Technology Innovation and Application, BWTAC 2024

作者： Zhang, Jinquan Wang, Chongbo Yang, Rendong Tang, Yuncan Ma, Yunshen Ni, Lina College of Computer Science and Engineering Shandong University of Science and Technology Qingdao266590 China Key Laboratory of the Ministry of Education for Embedded System and Service Computing Tongji University Shanghai201804 China

ISBN: (纸本)9789819794119

Federated Learning (FL), as one of the effective methods to solve the problem of medical data silos, can promote mutual cooperation among medical institutions under the premise of safeguarding the privacy and security of medical data, and effectively promote the development of intelligent medical treatment. However, the characteristics of medical data with multiple sources and heterogeneity bring certain challenges to the application of federated learning in the medical field. Due to the wide range of medical data sources, some medical institutions are unable to complete the federated learning training task on time due to the limitation of device performance, which leads to the failure of federated learning training. To address the problem of heterogeneity of computing devices among medical institutions that may cause some clients to fall behind, we propose an Adaptive Federated Learning based on Device Performance (AFedDP) method. Specifically, we design an adaptive local model training method to reduce the computational burden of medical institutions with poor equipment performance in the local model training process by adaptively constructing a local training model. Additionally, we propose a hierarchical model aggregation algorithm based on dynamic weight. This allows the heterogeneous adaptive local model to actively participate in the global model aggregation, making full use of the local data of each medical institution and improving the performance of the global model. Finally, we design an aggregation weight update algorithm based on momentum optimization to enhance the stability of global model aggregation. This algorithm updates contribution and aggregation weight with the idea of momentum optimization. Simulation experiments demonstrate that AFedDP achieves higher accuracy compared to the baseline algorithm. AFedDP can better address the issue of some devices being unable to participate in federated training properly. © The Author(s), under exclusive lic

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Convex Hull-based Algebraic Constraint for Visual Quadric SLAM

arXiv

引用

arXiv 2025年

作者： Yu, Xiaolong Zhao, Junqiao Song, Shuangfu Zhu, Zhongyang Yuan, Zihan Ye, Chen Feng, Tiantian School of Computer Science and Technology Tongji University Shanghai China The MOE Key Lab of Embedded System and Service Computing Tongji University Shanghai China Institute of Intelligent Vehicles Tongji University Shanghai China School of Surveying and Geo-Informatics Tongji University Shanghai China

Using Quadrics as the object representation has the benefits of both generality and closed-form projection derivation between image and world spaces. Although numerous constraints have been proposed for dual quadric reconstruction, we found that many of them are imprecise and provide minimal improvements to localization. After scrutinizing the existing constraints, we introduce a concise yet more precise convex hull-based algebraic constraint for object landmarks, which is applied to object reconstruction, frontend pose estimation, and backend bundle adjustment. This constraint is designed to fully leverage precise semantic segmentation, effectively mitigating mismatches between complex-shaped object contours and dual quadrics. Experiments on public datasets demonstrate that our approach is applicable to both monocular and RGB-D SLAM and achieves improved object mapping and localization than existing quadric SLAM methods. The implementation of our method is available at https://***/tievtongji/convexhull-based-algebraic-constraint. © 2025, CC BY.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Behavioral Unicity: On the Limits of Anonymized Social Behavior Metadata 1st

Behavioral Unicity: On the Limits of Anonymized Social Beh...

引用

1st Conference on Blockchain and Web3.0 Technology Innovation and Application, BWTAC 2024

作者： Tang, Hao Wang, Cheng Tongji University Shanghai201804 China The Key Laboratory of Embedded System and Service Computing Ministry of Education Shanghai201804 China Shanghai Artificial Intelligence Laboratory Shanghai200030 China Shanghai Institute of Intelligent Science and Technology Tongji University Shanghai200092 China

ISBN: (纸本)9789819794119

Social behavior metadata has altered business models and human lifestyles. However, the inclusion of personal information in social behavior metadata poses risks of identity exposure for users. Under the requirements of modern data management laws, service providers now employ anonymization techniques when collecting user behavior metadata, which aims to break the link between personal identifiers and behavioral attributes for privacy protection. This paper conducts experiments on an open-source dataset of anonymized social behavior metadata and finds that even with privacy protection techniques such as data generalization, there remains a high risk of reidentification. Specifically, four social behavior metadata records are sufficient to reidentify users from the anonymized social behavior metadata, achieving a unicity of 90.10%. This indicates that the current widespread collection of social behavior metadata brings a high risk of privacy breaches. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Anonymity

来源：评论

学校读者我要写书评

暂无评论

Joint DNN Partition and Thread Allocation Optimization for Energy-Harvesting MEC systems

Joint DNN Partition and Thread Allocation Optimization for E...

引用

Design, Automation and Test in Europe Conference and Exhibition

作者： Yizhou Shi Liying Li Yue Zeng Peijin Cong Junlong Zhou School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing China National Key Laboratory for Novel Software Technology Nanjing University Nanjing China National Mobile Communications Research Laboratory Southeast University Nanjing China Key Laboratory of Embedded System and Service Computing (Tongji University) Ministry of Education Shanghai China

ISBN: (数字)9783982674100

ISBN: (纸本)9798331534646

Deep neural networks (DNNs) have demonstrated exceptional performance, leading to diverse applications across various mobile devices (MDs). Considering factors like portability and environmental sustainability, an increasing number of MDs are adopting energy harvesting (EH) techniques for power supply. However, the computational intensity of DNNs presents significant challenges for their deployment on these resource-constrained devices. Existing approaches often employ DNN partition or offloading to mitigate the time and energy consumption associated with running DNNs on MDs. Nonetheless, existing methods frequently fall short in accurately modeling the execution time of DNNs, and do not consider to use thread allocation for further latency and energy consumption optimization. To solve these problems, we propose a dynamic DNN partition and thread allocation method to optimize the latency and energy consumption of running DNNs on EH-enabled MDs. Specifically, we first investigate the relationship between DNN inference latency and allocated threads and establish an accurate DNN latency prediction model. Based on the prediction model, a DRL-based DNN partition (DDP) algorithm is designed to find the optimal partitions for DNNs. A thread allocation (TA) algorithm is proposed to reduce the inference latency. Experimental results from our test-bed platform demonstrate that compared to four benchmarking methods, our scheme can reduce DNN inference latency and energy consumption by up to 37.3% and 38.5%.

关键词： Energy consumption Instruction sets Heuristic algorithms Artificial neural networks Predictive models Prediction algorithms Inference algorithms Partitioning algorithms Resource management Optimization

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

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

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：