检索结果-内蒙古大学图书馆

Few-shot transfer learning for wearable IMU-based human activity recognition

Neural Computing and Applications 2024年第18期36卷 10811-10823页

作者： Ganesha, H.S. Gupta, Rinki Gupta, Sindhu Hak Rajan, Sreeraman The Department of Electronics and Communication Engineering Amity University UP Noida201313 India The Amity Centre for Artificial Intelligence Amity University UP Noida201313 India The Department of Systems and Computer Engineering Carleton University Ottawa Canada

Deep learning has proven to be highly effective for human activity recognition (HAR) when large amount of labelled data is available for the target task. However, training a deep learning model to generalize well on a new task with just-few observations of labelled data is an active area of research. In this paper, a novel few-shot transfer learning (FSTL) approach is proposed for classification of human activities using just few instances (shots) of the data obtained from a wearable system assembled to collect inertial sensor data for different human activities, performed by two users. First, a deep learning model is trained on a large publicly available HAR dataset. The model parameters of such a model are then fine-tuned using the Reptile algorithm to determine the optimal initial parameter set using which, the model will classify activities with just few-shots of data from the target task. The proposed FSTL approach yields an average classification accuracy of 74.86 ± 0.71% and 79.20 ± 1.05% for 3-way, 5-shot classification of new activities performed by a single user and same set of activities performed by a new user, respectively. When the pre-trained weights are used as the initial weights in the Reptile algorithm, the generalization ability of the model improves by about 10% for 3-way, 5-shot classification as compared to using few-shot learning without parameter transfer. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Inertial navigation systems

来源：评论

学校读者我要写书评

暂无评论

Importance of Realistic Considerations of Time Synchronization for TAS in TSN Networks

Importance of Realistic Considerations of Time Synchronizati...

引用

2023 International Symposium on Networks, computers and Communications, ISNCC 2023

作者： Ghosh, Aviroop Kunz, Thomas Yousefi, Saleh Carleton University Department of Systems and Computer Engineering Ottawa Canada Urmia University Department of Computer Engineering Urmia Iran

ISBN: (纸本)9798350335590

Industry 4.0 applications require communication networks with deterministic and bounded latency. Ethernet, augmented with the IEEE Time-Sensitive Networking (TSN) standards, can meet these requirements. One of the TSN standards, Time Aware Shaper (TAS) is ideal for periodic time-sensitive traffic in wired TSN networks. However, it relies on network-wide synchronization to a common reference time. Existing solutions for developing TAS schedules assume perfect synchronization or consider a worst-case error. This study highlights the need for using network-derived time synchronization parameters and errors to optimize TAS schedules. It raises the importance of accurate time synchronization information and proposes a novel methodology for efficient band-width utilization while ensuring deadline requirements are met for time-sensitive traffic in realistic networks. The effectiveness of the proposed framework is validated by simulating a TSN network in OMNeT++ and comparing the devised TAS schedules with schedules which use existing assumptions. © 2023 IEEE.

关键词： Ethernet

来源：评论

学校读者我要写书评

暂无评论

DEVS AS A METHOD TO MODEL AND SIMULATE COMBINATORIAL DOUBLE AUCTIONS FOR E-PROCUREMENT

DEVS AS A METHOD TO MODEL AND SIMULATE COMBINATORIAL DOUBLE ...

引用

2024 Winter Simulation Conference, WSC 2024

作者： De Antón, Juan Ruiz-Martin, Cristina Villafáñez, Félix Poza, David Department of Business Organization INSISOC University of Valladolid Valladolid Spain Department of Systems and Computer Engineering Carleton University Ottawa Canada

ISBN: (纸本)9798331534202

The surge in electronic procurement is fostering the proliferation of electronic marketplaces and advanced auctions as primary coordination mechanisms. Among these, combinatorial and double auctions are gaining traction in the procurement sector. However, prevalent implementations often assume participants to be perfectly rational, adhering to predefined behaviors within the auction model. These centralized models, while prevalent, fail to capture the intricate dynamics of real auction environments adequately. Consequently, there is a growing recognition of the necessity for decentralized models within an agent-based framework to simulate such auctions authentically. The contribution of this work is the application of the DEVS formalism to develop a decentralized model for a combinatorial iterative double auction to address the limitations of centralized implementations. The model is formally defined, and a case study is presented to verify it against its centralized version. This is the first step toward accommodating agents with varied behavioral patterns within auction simulations. © 2024 IEEE.

关键词： Decentralized systems

来源：评论

学校读者我要写书评

暂无评论

Sources of error during inertial sensing of human movement: a critical review of the fundamentals

Sources of error during inertial sensing of human movement: ...

引用

2023 IEEE International Symposium on Medical Measurements and Applications, MeMeA 2023

作者： Beange, Kristen H.E. Chan, Adrian D.C. Graham, Ryan B. Carleton University Department of Systems and Computer Engineering Ottawa Canada University of Ottawa School of Human Kinetics Ottawa Canada

ISBN: (纸本)9781665493840

Inertial assessments of human movement have potential to support diagnosis and treatment of neuromuscular disorders in healthcare settings. Despite the potential advantages, uptake and acceptance by healthcare professionals are still a challenge, as inertial measurement units are prone to measurement errors due to inherent limitations with the technology. As such, full exploitation is limited to a small group of highly qualified personnel. For usage to be more ubiquitous, standard practices for acquiring high-quality data are required and should include methods for error avoidance, detection, identification, quantification, and mitigation. In this paper, a critical review of sources of error was conducted, from which a taxonomic error classification framework was developed. From this review, it has become apparent which sources of error carry the highest risk for impacting data quality. Methods for error mitigation have been identified, along with limitations and areas for improvement. This framework is intended to serve as a useful reference for both proficient and non-proficient users to ensure all sources of error are considered when developing and interpreting IMU-based assessments. It also provides a foundation for developing standard practices to help users efficiently and reliably acquire high-quality data, which is imperative for uptake and acceptance in healthcare settings. © 2023 IEEE.

关键词： Measurement errors

来源：评论

学校读者我要写书评

暂无评论

Event-Triggered Fixed-Time Consensus of Second-Order Nonlinear Multi-Agent systems with Delay and Switching Topologies

引用

Journal of Shanghai Jiaotong university(Science) 2024年第4期29卷 625-639页

作者： XING Youjing GAO Jinfeng LIU Xiaoping WU Ping School of Information Science and Engineering Zhejiang Sci-Tech UniversityHangzhou310018China Department of Systems and Computer Engineering Carleton UniversityOttawaK1S 5B6Canada

To address fixed-time consensus problems of a class of leader-follower second-order nonlinear multi-agent systems with uncertain external disturbances,the event-triggered fixed-time consensus protocol is ***,the virtual velocity is designed based on the backstepping control method to achieve the system consensus and the bound on convergence time only depending on the system ***,an event-triggered mechanism is presented to solve the problem of frequent communication between agents,and triggered condition based on state information is given for each *** is available to save communication resources,and the Zeno behaviors are ***,the delay and switching topologies of the system are also ***,the system stabilization is analyzed by Lyapunov stability ***,simulation results demonstrate the validity of the presented method.

关键词： event-triggered mechanism fixed-time consensus multi-agent systems switching topologies

来源：评论

学校读者我要写书评

暂无评论

A deep learning based misbehavior classification scheme for intrusion detection in cooperative intelligent transportation systems

引用

Digital Communications and Networks 2023年第5期9卷 1113-1122页

作者： Tejasvi Alladi Varun Kohli Vinay Chamola F.Richard Yu Department of Computer Science and Information Systems BITS-PilaniPilani Campus333031India Department of Electrical and Electronics Engineering&APPCAIR BITS-PilaniPilani Campus333031India Department of Systems and Computer Engineering Carleton UniversityOttawaONK1S 5B6Canada

With the rise of the Internet of Vehicles(IoV)and the number of connected vehicles increasing on the roads,Cooperative Intelligent Transportation systems(C-ITSs)have become an important area of *** the number of Vehicle to Vehicle(V2V)and Vehicle to Interface(V2I)communication links increases,the amount of data received and processed in the network also *** addition,networking interfaces need to be made more secure for which existing cryptography-based security schemes may not be ***,there is a need to augment them with intelligent network intrusion detection *** machine learning-based intrusion detection and anomaly detection techniques for vehicular networks have been proposed in recent ***,given the expected large network size,there is a necessity for extensive data processing for use in such anomaly detection *** learning solutions are lucrative options as they remove the necessity for feature ***,with the amount of vehicular network traffic increasing at an unprecedented rate in the C-ITS scenario,the need for deep learning-based techniques is all the more *** work presents three deep learning-based misbehavior classification schemes for intrusion detection in IoV networks using Long Short Term Memory(LSTM)and Convolutional Neural Networks(CNNs).The proposed Deep Learning Classification Engines(DCLE)comprise of single or multi-step classification done by deep learning models that are deployed on the vehicular edge *** data received by the Road Side Units(RSUs)is pre-processed and forwarded to the edge server for classifications following the three classification schemes proposed in this *** proposed classifiers identify 18 different vehicular behavior types,the F1-scores ranging from 95.58%to 96.75%,much higher than the existing *** running the classifiers on testbeds emulating edge servers,the prediction performance and prediction time comparison of

关键词： Vehicular Ad-hoc Networks(VANETs) Intelligent Transportation systems(ITS) Artificial Intelligence(AI) Deep Learning Internet of Things(IoT)

来源：评论

学校读者我要写书评

暂无评论

Survey of Commercial Wi-Fi Devices' Energy Detection Mechanism

Survey of Commercial Wi-Fi Devices' Energy Detection Mechani...

引用

2024 IEEE International Conference on Consumer Electronics, ICCE 2024

作者： Bailey, Jeffrey Liu, Stone Lambadaris, Ioannis Max, Sebastian Sugirtharaj, David Carleton University Department of Systems and Computer Engineering Ottawa Canada Radio Network Research Ericsson Germany Ericsson Sweden

ISBN: (纸本)9798350324136

This paper presents an experimental survey of the capabilities of commercial IEEE 802.11 (Wi-Fi) devices to detect the state of the wireless medium. Wi-Fi devices use a Clear Channel Assessment (CCA) mechanism to determine if the wireless medium is occupied. By subjecting Wi-Fi devices to Gaussian interference signals transmitted through a Universal Software Radio Peripheral (USRP), we assess their CCA behavior. Moreover, we perform a deeper investigation of the Ericsson AP6321 Wi-Fi Access Point prompted by privileged access to its internal operations. Our findings highlight variations between different Wi-Fi devices' CCA mechanisms and reveal floating CCA thresholds. Since CCA forms the fundamental basis for Wi-Fi operation, this study provides insights into coexistence of Wi-Fi with itself and dissimilar technologies. Because future Narrow Band Frequency Hopping (NB FH) systems are a potential technology that commercial Wi-Fi devices need to coexist with, we provide an extended survey of the capability of Wi-Fi devices to detect NB FH signals in unlicensed bands. This survey demonstrates that rapidly changing, on-off signalling does not have a disproportionate impact on Wi-Fi performance. © 2024 IEEE.

关键词： Adaptive Noise Floor Clear Channel Assessment (CCA) Coexistence Dynamic On-Off (DOF) Interference Energy Detection Threshold (EDT) Narrow Band Frequency Hopping (NB FH)

来源：评论

学校读者我要写书评

暂无评论

Deep Learning-Based Anomaly Detection in 5G Cellular Networks 15

Deep Learning-Based Anomaly Detection in 5G Cellular Network...

引用

15th International Conference on Information, Intelligence, systems and Applications, IISA 2024

作者： Farooq, Umar Hameed, Aroosa Leivadeas, Aris Lambadaris, Ioannis École de Technologie Supérieure Department of Software and It Engineering Montreal Canada Carleton University Department of Systems and Computer Engineering Ottawa Canada

ISBN: (纸本)9798350368833

The telecommunication industry saw a dramatic shift with the introduction of 5G technologies. This new cellular generation brought lightening speed, massive capacity, and better connectivity. Despite the many introduced opportunities, several challenges emerge at the same time, especially in the field of performance assurance. An example of such challenges is the rapid identification of network anomalies, which is critical for maintaining network performance and ensuring user satisfaction. Traditional techniques of detecting anomalies have fallen short given the unique operating requirements of 5G networks. To address this issue, this paper presents a novel technique applying deep learning for the detection of cellular network anomalies. Specifically, we leverage and exploit the capabilities of several Long-Short-Term-Memory (LSTM) and Artificial Neural Networks (ANN) flavors, such as LSTM with ANN and Bidirection-LSTM (BiLSTM) with ANN, to excel at identifying network anomalies for the preservation and efficiency of 5G networks. Our work includes extensive experimentation and the attained results show that our proposed models are overwhelmingly adapting to preserving the integrity of the network in the fast-paced, ever-changing realm of cellular networks. © 2024 IEEE.

关键词： 5G mobile communication systems

来源：评论

学校读者我要写书评

暂无评论

Security-Oriented Network Intent Placement using Particle Swarm Optimization 7

Security-Oriented Network Intent Placement using Particle Sw...

引用

7th Cyber Security in Networking Conference, CSNet 2023

作者： Landeau, Gabriel Avgeris, Marios Leivadeas, Aris Lambadaris, Ioannis Department of Software and It Engineering Montréal Canada Carleton University Department of Systems and Computer Engineering Ottawa Canada

ISBN: (纸本)9798350342871

As the network infrastructure grows, its configuration and service provisioning become a tedious process. Accordingly, new paradigms have emerged, such as the Intent-Based Networking (IBN), that envision the automation of the network configuration, while minimizing the human intervention. Specifically, IBN allows users to interact with the network through high-level and declarative requests, called intents, which later can be translated into low-level configurations. IBN can entail different scopes and target network infrastructures, while being domain specific, which can create several challenges in terms of the final activation of the requested intents. To this end, in this paper, we mainly focus on intents that are expressing security and Quality of Service (QoS) network services demands that can be translated into Service Function Chains (SFC) and automatically deployed over a campus network. Our work and depending on the security level expressed in the intent, tries to optimally decide the level of multi-tenancy or complete segregation of the users' services that can be achieved, while satisfying the network provider's objectives. In particular, an artificial intelligence inspired algorithm called Particle Swarm Optimization (PSO) is modeled that automatically tries to find the best placement of the intents, while satisfying the security and QoS requirements of the users issuing the intents. © 2023 IEEE.

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

Assessing Driver Engagement Through Machine Learning Classification of Physiological Measures

Assessing Driver Engagement Through Machine Learning Classif...

引用

2023 IEEE International Symposium on Medical Measurements and Applications, MeMeA 2023

作者： Lochbihler, Aidan Wallace, Bruce Benthem, Kathleen Van Herdman, Chris Sloan, Will Brightman, Kirsten Goheen, Josh Knoefel, Frank Marshall, Shawn Carleton University Systems and Computer Engineering Ottawa Canada Carleton University Bruyère Research Institute Sam 3 Innovation Hub Systems and Computer Engineering Ottawa Canada Carleton University Department of Cognitive Science Ottawa Canada Carleton University Bruyère Research Institute Elizabeth Bruyère Hospital Systems and Computer Engineering Ottawa Canada Carleton University Bruyère Research Institute Elizabeth Bruyère Hospital Ottawa Hospital Ottawa Canada

ISBN: (纸本)9781665493840

The assessment to determine if drivers are engaged and supervising autonomous and semi-autonomous vehicles (AV) is becoming an increasingly important task. AVs are becoming more prevalent on the road and adequate driver engagement is a must for safe operation. The full adoption of level 5 vehicles will likely take many years as they do not currently exist for consumer purchase and before this happens the roads will see a mix of level 0-4 vehicles. During this time humans will still be responsible for taking control of the vehicle if a hazardous scenario occurs and the AV does not know how to maneuver around. To have a safe handover from AV to human during these situations the driver must maintain a level of engagement even while the AV is driving. To do this physiological sensors can be used to measure signals such as heart rate and respiration rate, which are known indicators of a driver's engagement. This paper exposes drivers to non-surprise and surprise driving scenarios to assess if attentive drivers can be identified from physiological changes for manual and AV driving. Machine learning (ML) is used to understand the patterns of physiological signals and classify when a driver is engaged during a surprise scenario. Finally, the ML models show a successful ability to classify engaged versus non-engaged drivers with a 73.3% accuracy in manual driving scenarios, 86.7% accuracy in AV, and 70.0% accuracy when the data from both driving scenarios are combined showing the model's ability to generalize. © 2023 IEEE.

关键词： Autonomous vehicles

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：