检索结果-内蒙古大学图书馆

arXiv 2020年

作者： Chen, Po-Yu Bhatia, Laksh Kolcun, Roman Boyle, David McCann, Julie A. Department of Computing Imperial College London Dyson School of Design Engineering Imperial College London

LoRaWAN is one of the leading Low Power Wide Area Network (LPWAN) architectures. It was originally designed for systems consisting of static sensor or Internet of Things (IoT) devices and static gateways. It was recently updated to introduce new features such as nano-second timestamps which open up applications to enable LoRaWAN to be adopted for mobile device tracking and localisation. In such mobile scenarios, devices could temporarily lose communication with the gateways because of interference from obstacles or deep fading, causing throughput reduction and delays in data transmission. To overcome this problem, we propose a new data forwarding scheme. Instead of holding the data until the next contact with gateways, devices can forward their data to nearby devices that have a higher probability of being in contact with gateways. We propose a new network metric called Real-Time Contact-Aware Expected Transmission Count (RCA-ETX) to model this contact probability in real-time. Without making any assumption on mobility models, this metric exploits data transmission delays to model complex device mobility. We also extend RCA-ETX with a throughput-optimal stochastic backpressure routing scheme and propose Real-Time Opportunistic Backpressure Collection (ROBC), a protocol to counter the stochastic behaviours resulting from the dynamics associated with mobility. To apply our approaches seamlessly to LoRaWAN-enabled devices, we further propose two new LaRaWAN classes, namely Modified Class-C and Queue-based Class-A. Both of them are compatible with LoRaWAN Class-A devices. Our data-driven experiments, based on the London bus network, show that our approaches can reduce data transmission delays up to 25% and provide a 53% throughput improvement in data transfer performance. Copyright © 2020, The Authors. All rights reserved.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

When Adversarial Example Attacks Meet Vertical Federated Learning

When Adversarial Example Attacks Meet Vertical Federated Lea...

引用

Symposia and Workshops on Ubiquitous, Autonomic and Trusted computing, UIC-ATC

作者： Dan Meng Zhihui Fu Chao Kong Yue Qi Guitao Cao OPPO Research Institute Shenzhen China School of Computer and Information Anhui Polytechnic University Wuhu China Shanghai Key Laboratory of Trustworthy Computing East China Normal University Shanghai China MoE Engineering Research Center of SH/HW Co-design Technology and Application ECNU Shanghai China

Federated learning (FL) is proposed to enable efficient machine learning while protecting the privacy of user data. In non-federated scenarios, machine learning models are vulnerable to adversarial attacks. The attacker can construct elaborate adversarial examples, which can greatly reduce the performance of the victim model. It is worth exploring how FL models differ from traditional models in face of complex adversarial attacks. We are also curious about the robustness of different FL modeling methods against adversarial attacks, providing suggestions for selecting suitable FL models in practical applications. Consequently, we perform adversarial example attacks on vertical federated learning and design experiments to study how vertical federated models perform under adversarial example attacks. The simulation results show that federated models are more robust than non-federated models. Besides, in vertical federation learning, decision trees and XGBoost are much more robust than logistic regression and neural networks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Assessing Microclimate Impacts of Neighborhood Redesign in a Desert Urban Climate Using ENVI-Met and MaRTy

SSRN

引用

SSRN 2022年

作者： Crank, Peter J. Middel, Ariane Coseo, Paul J. Sailor, David J. College of Integrative Studies Singapore Management University Singapore Singapore School of Arts Media and Engineering Arizona State University United States School of Computing and Augmented Intelligence Arizona State University United States The Design School Arizona State University United States School of Geographical Sciences and Urban Planning Arizona State University United States Urban Climate Research Center Arizona State University United States

Municipalities often consider heat mitigation strategies to address urban overheating, but the location of implementation rarely is co-located with the communities that are carrying the majority of the heat burden in the city. The City of Phoenix, with a history of environmental racism, is redeveloping a public housing community with a focus on urban cooling as a desired outcome. This research uses in situ measurements (including the mobile micro-meteorological measurement cart, MaRTy) and ENVI-met microscale modeling of the neighborhood to assess air temperature cooling capabilities of the planned redesigns to the neighborhood. After validating the ENVI-met model of the current neighborhood with fixed and mobile measurements with an RMSE 0.94 and d > 0.9, respectively, analysis of the planned urban design shows some cool spots connected to new shade and vegetated corridors with air temperature cooling magnitudes as high as 1.5 °C. Yet, some exposed and building-adjacent areas were identified as potential hot spots in the planned neighborhood. These hotspots underscore the importance of continued collaboration among the City, researchers, and the community to address the needs of the community for the creation of healthier urban environments. © 2022, The Authors. All rights reserved.

关键词： Atmospheric temperature

来源：评论

学校读者我要写书评

暂无评论

Using Weak Time-Delay and Transfer Learning to Improve Neural Network Performance in the Determination of Coagulant Dosing for Drinking Water Plants

SSRN

引用

SSRN 2023年

作者： Li, Si Zhu, Guocheng Fang, Haiquan Li, Xiaoshang Yuan, Cheng Yuan, Fang Wang, Jian Hursthouse, Andrew S. College of Civil Engineering Hunan University of Science and Technology Hunan Xiangtan411201 China Xiangtan Middle Ring Water Business Limited Corporation Hunan Xiangtan411201 China Tsinghua University Guangdong Provincial Engineering Technology Research Center Urban Water C Tsinghua Shenzhen Int Grad School Shenzhen518055 China Guangdong Provincial Engineering and Technology Research Center for Water Affairs Big Data a nd Water Ecology Shenzhen Water Planning & Design Institute Co. Ltd. Shenzhen518001 China School of Computing Engineering & Physical Sciences University of the West of Scotland PaisleyPA1 2BE United Kingdom

Coagulation is the most significant unit operation in drinking water treatment plants, which involves the precise control of coagulant dose, to minimize chemical expenditure. However, due to the significant time lag and external influences on the coagulation process, it can be difficult to build an appropriate coagulant dose prediction model. In this study, we proposed the concept of weak time delay and transfer learning to train the model and explain the sensitivity of factors affecting the prediction of coagulant dose, to improve the neural network model. The results showed that with the weak time-delay data including raw water and settling water quality parameters of one-hour delayed test, the model result improved with a more sensitive coagulant dose prediction. The sensitive factors included raw water turbidity, flow rate, temperature, total dissolved solids and settling water turbidity. Using these parameters, we were able to construct an effective model for prediction of coagulant dose, with a R2 correlation coefficient up to 0.72. The transfer learning technique successfully introduced the structure and parameters of the laboratory model for the training of the drinking water plant model. The results were consistent with the genetic algorithm, but its training speed increased by hundreds of times, and could also overcome the disadvantage of over fitting. Overall, the findings from this study provide new methods and technical support for the establishment of coagulation dose, neural network model. © 2023, The Authors. All rights reserved.

关键词： Time delay

来源：评论

学校读者我要写书评

暂无评论

3DLG-Detector: 3D Object Detection via Simultaneous Local-Global Feature Learning

arXiv

引用

arXiv 2022年

作者： Chen, Baian Nan, Liangliang Xie, Haoran Lu, Dening Wang, Fu Lee Wei, Mingqiang The School of Computer Science and Technology Nanjing University of Aeronautics and Astronautics Nanjing China The Urban Data Science Section Delft University of Technology Delft Netherlands The Department of Computing and Decision Sciences Lingnan University Hong Kong The Department of Systems Design Engineering University of Waterloo Waterloo Canada The School of Science and Technology Hong Kong Metropolitan University Hong Kong

Capturing both local and global features of irregular point clouds is essential to 3D object detection (3OD). However, mainstream 3D detectors, e.g., VoteNet and its variants, either abandon considerable local features during pooling operations or ignore many global features in the whole scene context. This paper explores new modules to simultaneously learn local-global features of scene point clouds that serve 3OD positively. To this end, we propose an effective 3OD network via simultaneous local-global feature learning (dubbed 3DLG-Detector). 3DLG-Detector has two key contributions. First, it develops a Dynamic Points Interaction (DPI) module that preserves effective local features during pooling. Besides, DPI is detachable and can be incorporated into existing 3OD networks to boost their performance. Second, it develops a Global Context Aggregation module to aggregate multi-scale features from different layers of the encoder to achieve scene context-awareness. Our method shows improvements over thirteen competitors in terms of detection accuracy and robustness on both the SUN RGB-D and ScanNet datasets. Source code will be available upon publication. © 2022, CC BY.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Research on Post-evaluation of Pavement Performance of Municipal Roads Based on FWD

Research on Post-evaluation of Pavement Performance of Munic...

引用

2022世界交通运输大会(WTC2022)

作者： Chao-liang Gou Xiao-lei Nong Zi-ye Zhang Zhan-chun Li Zhong-lu Cao Qin-shou Huang Jin-ping Chen Kai-wei Zeng Yong-jun Meng School of civil engineering Guangxi University CCCC-FHDI Engineering Co. Ltd.Guangzhou Guangxi Road Construction Engineering Group Co.LTO. CCCC Tianjin Port Engineering Institute Co.Ltd. Guangxi Communications Design Group Co. Ltd. Key Laboratory of engineering disaster prevention and structural safety Ministry of education Guangxi special geological Highway Safety Engineering Technology Research Center National High-performance Computing Center Nanning Branch Nanning

In order to study the feasibility of using FWD to post-evaluate the structural performance of municipal road pavement,the main factors affecting the bending settlement and modulus results of the pavement were analyzed,and the inverse calculation of FWD test results was performed by using the algorithm program prepared by VB6.0,and thePavement Structure Performance Index(PSPI)was used as evaluation indexes to post-evaluate the municipal *** results show that the modulus of the pavement structural layer has the greatest influence on the bending settlement,and the accuracy of the bending settlement should be guaranteed above 1/1000 mm to ensure the accuracy of the modulus results,and the PSPI is used as evaluation indexes to construct a logical framework model of the pavement structural performance of municipal roads can be used as guidelines for the actual maintenance and repair of roads.

关键词： Municipal roads FWD Post-evaluation Structural performance Dynamic modulus

来源：评论

学校读者我要写书评

暂无评论

DCET-Net: Dual-Stream Convolution Expanded Transformer for Breast Cancer Histopathological Image Classification

DCET-Net: Dual-Stream Convolution Expanded Transformer for B...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Ying Zou Shannan Chen Qiule Sun Bin Liu Jianxin Zhang Key Lab of Advanced Design and Intelligent Computing Ministry of Education Dalian University Dalian China School of Information and Communication Engineering Dalian University of Technology Dalian China International School of Information Science and Engineering (DUT-RUISE) Dalian University of Technology Dalian China School of Computer Science and Engineering Dalian Minzu University Dalian China

ISBN: (纸本)9781665429825

Researches on breast cancer histopathological image classification have achieved a great breakthrough using deep backbones of Convolutional Neural Networks (CNNs) in recent years. However, due to the inductive bias of locality, CNNs are unable to effectively extract the global feature information of breast cancer histopathological images, limiting the improvement of the classification results. To overcome this shortcoming, this paper reasonably introduces an extra backbone stream of a pure transformer, which consists of a self-attention mechanism to capture global receptive fields of histopathological images, thereby compensating the locality characteristic of CNNs backbone. Based on two backbone streams of CNN and transformer, a dual-stream network called DCET-Net is proposed, which considers local features and global ones simultaneously, and progressively combines them from these two streams to form the final representations for classification. DCET-Net is extensively evaluated on the representative BreakHis histopathological image dataset, and experimental results demonstrate that it is highly competitive with the state-of-the-art CNN methods in breast cancer histopathological image classification task.

关键词： Limiting Biological system modeling Streaming media Transformers Feature extraction Breast cancer Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Lithium-ion battery degradation: How to model it

arXiv

引用

arXiv 2021年

作者： O’Kane, Simon E.J. Ai, Weilong Madabattula, Ganesh Alvarez, Diego Alonso Edge, Jacqueline Sophie Wu, Billy Offer, Gregory J. Marinescu, Monica Department of Mechanical Engineering Imperial College London United Kingdom Dyson School of Design Engineering Imperial College London United Kingdom Research Computing Service ICT Imperial College London United Kingdom The Faraday Institution United Kingdom

Predicting lithium-ion battery degradation is worth billions to the global automotive, aviation and energy storage industries, to improve performance and safety and reduce warranty liabilities. However, there are no reliable models and developers must resort to expensive and time consuming experiments, and only a few can afford to test under every realistic combination of operating conditions. In this paper, we report the first published attempt to directly couple more than two degradation mechanisms in the negative electrode, and map different pathways through the complicated path dependent and nonlinear degradation space. Four degradation mechanisms are coupled in PyBaMM, an open source modelling environment uniquely developed to allow new physics to be implemented and explored quickly and easily. We have run 30 simulations of 1000 cycles, which would have cost $ 300000 experimentally. Crucially it is possible to see ‘inside’ the model and observe the consequences of the different patterns of degradation, such as loss of lithium inventory and loss of active material. For the same cell we have already uncovered five different pathways that can result in end-of-life, depending on how the cell is used. Such information would enable a product designer to either extend life or predict life based upon the usage pattern. However, parameterization of the degradation models remains as a major challenge, and requires the attention of the international battery community. Copyright © 2021, The Authors. All rights reserved.

关键词： Degradation

来源：评论

学校读者我要写书评

暂无评论

Preliminary study on the spread of air-borne pollutants in urban environment: a CFD simulation approach

引用

Scientific Reports 2025年第1期15卷 1-19页

作者： Ahmad, Fatma Majumder, Debjit Ranjit, Rabs Gupta, Aman Manhart, Michael Chair of Computing in Civil Engineering and Geoinformation Systems and Geodetic Institute RWTH Aachen University Aachen Germany School of Engineering and Design Technical University of Munich Munich Germany Institute of Geography School of Geosciences University of Edinburgh Edinburgh United Kingdom Department of Architecture and Planning Indian Institute of Engineering Science and Technology Shibpur India

The spreading of pollutants within urban areas, particularly from traffic emissions, poses a significant health risk. Computational Fluid Dynamics (CFD) has emerged as a key tool in understanding how pollutants spread within a city. In particular, the large-eddy simulation (LES) approach allows us to capture the complex time-dependent behaviour of the 3D flow field due to buildings in a dense urban environment. This work utilizes the CFD tool MGLET (Multi Grid Large Eddy Turbulence) to model the transport of pollutants within a selected domain in Munich city. MGLET offers a feature of simulating transport of multiple passive scalar quantities simultaneously. This facilitates the individual analysis of emissions from each major street in the domain of interest, providing detailed insights into their respective impacts. Additionally, MGLET utilizes the Immersed Boundary Method to resolve 3D building geometries, removing the need to generate body-fitted grids, which tends to be highly time-consuming. The streets are defined by area sources, and the emission rates for each street are defined by the average traffic flow rate. This high-fidelity approach offers a detailed analysis, allowing us to identify local features such as recirculation zones in street canyons and pinpoint the streets that contribute most significantly to pollution at specific locations. The insights from CFD studies can empower policymakers to craft legislation tailored to local pollution control efforts, thus enhancing the quality of life in urban areas. Ultimately, the accurate prediction of pollutant concentration is critical, as it directly impacts the health and well-being of urban residents, highlighting the urgent need for effective pollution control measures. © The Author(s) 2025.

关键词： Air quality Computational fluid dynamics Large eddy simulation Wind simulation

来源：评论

学校读者我要写书评

暂无评论

t-PSI: Efficient Multi-party Private Set Intersection with Threshold

t-PSI: Efficient Multi-party Private Set Intersection with T...

引用

Symposia and Workshops on Ubiquitous, Autonomic and Trusted computing, UIC-ATC

作者： Dan Meng Zhihui Fu Chao Kong Yue Qi Guitao Cao OPPO Research Institute Shenzhen China School of Computer and Information Anhui Polytechnic University Wuhu China Shanghai Key Laboratory of Trustworthy Computing1 East China Normal University Shanghai China MoE Engineering Research Center of SH/HW Co-design Technology and Application ECNU Shanghai China

Multi-party privacy set intersection (MPSI) enables multiple parties to compute the intersection of their datasets without leaking data privacy. Among existing MPSI protocols, e.g., KMPRT-based protocols, oblivious evaluation of programmable pseudo-random function (OPPRF) or oblivious pseudo-random function (OPRF) is typically used to generate pseudo random numbers, requiring frequent online interactions. To mitigate such communication overheads, we propose the threshold Privacy Set Intersection (t-PSI) protocol for multi-party, leveraging the Shamir Secret Sharing (SSS) protocol and the masking construction mechanism. We propose two types of masking construction mechanisms: Polynomial and Garbled Bloom-filter. To the best of our knowledge, this is the first Secret Sharing-based MPSI protocol. With the significant reduction of inter-party communication, the t-PSI protocol is highly scalable and can handle datasets ranging from 2 6 to 2 20 . Further, with built-in fault tolerance, t-PSI works even if some parties went offline during protocol execution. As shown in the experiments, for five parties with datasets of 2 20 items each, the proposed t-PSI takes only 63 seconds, outperforming the state-of-the-art in [1].

关键词：

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：