检索结果-内蒙古大学图书馆

4th acm/ieee symposium on edge computing (SEC)

作者： Rausch, Thomas Raith, Philipp Pillai, Padmanabhan Dustdar, Schahram TU Wien Vienna Austria Intel Labs Vienna Austria

ISBN: (纸本)9781450367332

We present an end-to-end system for operating energy-aware cloudlets with a low-footprint cluster manager and an adaptive client-side load balancing approach. Our system is designed for small-scale high-density compute clusters that host stateless services and have stringent energy resource constraints. It features cluster and service management, runtime monitoring, adaptive load balancing and cluster reconfiguration policies. Furthermore, we present an experimentation and analytics system that allows coordinated execution of complex workload experiments to evaluate different operational strategies.

关键词： edge computing cloudlets energy-awareness load-balancing

来源：评论

学校读者我要写书评

暂无评论

Traffic Congestion Assessment Based on Street Level Data for On-edge Deployment 19

Traffic Congestion Assessment Based on Street Level Data for...

引用

4th acm/ieee symposium on edge computing (SEC)

作者： Li, Yuan Wang, Haili Buckles, Bill Univ North Texas Denton TX 76203 USA

ISBN: (纸本)9781450367332

With the increasing technological advancements and fields such as the industrial internet, autonomous driving, smart homes, intelligent transportation, and smart cities, all may benefit greatly from edge applications. These fields demand greater speed and efficiency for computing, network transmission, and user interaction in order to keep up with the growth. Stemming from the unique speed-up advantages that the edge computation model gives over traditional cloud computation, it is a suitable technology for deployment on the roads and streets, making edge a prime candidate to process traffic data. As previously mentioned, intelligent transportation, autonomous driving and smart cities all require intense computation to ensure they function correctly, and deploying edge on roads serves to address this issue. We explore the KITTI data set from the perspective of analyzing traffic congestion on complex roads. Our analysis shows: 1) based on object bounding boxes expressed within the data, each instantaneous situation can be separated into non-congestion and congestion;2) the mid-point distances between bounding boxes effectively determines the congestion level. To the best of our knowledge, this is the first proposed use of a non-continuous timestamped data set for traffic congestion detection.

关键词： edge computing spatial displacement CNN congestion appropriation

来源：评论

学校读者我要写书评

暂无评论

Learning-Based Microservice Placement and Migration for Multi-Access edge computing

引用

ieee TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT 2024年第2期21卷 1969-1982页

作者： Ray, Kaustabha Banerjee, Ansuman Narendra, Nanjangud C. Indian Stat Inst Adv Comp & Microelect Unit Kolkata 700108 India IBM Res Bengaluru India Ericsson Res Bengaluru 560048 India

In Multi-Access edge computing (MEC), a number of mechanisms exist to determine the optimal placement of monolithic service workflows. For applications designed as microservice workflow architectures, service placement schemes need to be revisited owing to the inherent interdependencies which exist between microservices. The dynamic environment, with stochastic user movement and service invocations, along with a large placement configuration space makes microservice placement in MEC a challenging task. Additionally, owing to user mobility, a placement scheme may need to be recalibrated, triggering service migrations to maintain the advantages offered by MEC. Existing microservice placement and migration schemes consider on-demand strategies. In this work, we take a different route and propose a Reinforcement Learning (RL) based proactive mechanism using a Learning Automata (LA) for microservice placement and migration that on one hand, keeps track of user mobility and resorts to migration when necessary, while on the other hand, keeps track of server residual capacities so that no server is overloaded. We use the San Francisco Taxi dataset to validate our approach. Experimental results show the effectiveness of our approach in comparison to other methods.

关键词： Microservice placement microservice migration reinforcement learning learning automata

来源：评论

学校读者我要写书评

暂无评论

Machine Learning Enabled Distributed Mobile edge computing Network 19

Machine Learning Enabled Distributed Mobile Edge Computing N...

引用

4th acm/ieee symposium on edge computing (SEC)

作者： Ma, Junchao Chang, Hao-Hsuan Fan, Pingzhi Liu, Lingjia Virginia Tech Bradley Dept Elect & Comp Engn Blacksburg VA 24061 USA Southwest Jiaotong Univ Key Lab Informat Coding & Transmiss Chengdu Peoples R China

ISBN: (纸本)9781450367332

In this work, we propose to establish a mobile edge computing (MEC) network that considers computation, caching and communication jointly. Depending on the demanding categories, users in the network are partitioned into computation-driven and cachingdriven users, both of which need memory resource to improve their quality of experiences (QoEs). Thus, a memory resource allocation problem is aroused to maximize the performance of the whole network. Due to the fact that the users' characterization plays an important role to the resource allocation scheme and with the help of machine learning techniques, we propose to study and predict the users' patterns by distributed learning methods which take the heterogeneity of base station type and users' mobility, etc into consideration. The proposed machine learning based distributed MEC system can maximize the efficiency of the network by optimizing the resource allocation scheme and perfectly predicting users' pattern.

关键词： Mobile edge computing

来源：评论

学校读者我要写书评

暂无评论

Poster: Virtual Reality Streaming at the edge: A Power Perspective 4

Poster: Virtual Reality Streaming at the Edge: A Power Persp...

引用

4th acm/ieee symposium on edge computing (SEC)

作者： Zhu, Zichen Jiang, Nan Guo, Tian Wei, Sheng Rutgers State Univ Piscataway NJ 08854 USA Univ Nebraska Lincoln NE USA Worcester Polytech Inst Worcester MA 01609 USA

ISBN: (纸本)9781450367332

This poster focuses on addressing the power consumption issues in 360-degree immersive video streaming on smart phones, an emerging virtual reality (VR) application in the consumer video market. We first conducted a power measurement study that indicates VR view generation as the major power consumption source. Then, we developed an edge-based immersive streaming system called edgeVR that offloads the power-consuming view generation operation from the smart phone to the edge. Through our preliminary evaluations using edgeVR, we identified the challenge of Motion-to-Photon latency associated with offloading. To reduce such latency, we propose a viewport prediction-based pre-rendering mechanism at the edge to ensure the quality of experience in the VR application.

关键词： Virtual reality

来源：评论

学校读者我要写书评

暂无评论

Task-Adaptive Incremental Learning for Intelligent edge Devices 19

Task-Adaptive Incremental Learning for Intelligent Edge Devi...

引用

4th acm/ieee symposium on edge computing (SEC)

作者： Qin, Zhuwei Yu, Fuxun Chen, Xiang George Mason Univ Fairfax VA 22030 USA

ISBN: (纸本)9781450367332

Convolutional Neural Networks (CNNs) are used for a wide range of image-related tasks such as image classification and object detection. However, a large pre-trained CNN model contains a lot of redundancy considering the task-specific edge applications. Also, the statically pre-trained model could not efficiently handle the dynamic data in the real-world application. The CNN training data and their label are collected in an incremental manner. To tackle the above two challenges, we proposed TeAM a task-adaptive incremental learning framework for CNNs in intelligent edge devices. Given a pre-trained large model, TeAM can configure it into any specialized model for dedicated edge applications. The specialized model can be quickly fine-tuned with local data to achieve very high accuracy. Also, with our global aggregation and incremental learning scheme, the specialized CNN models can be collaboratively aggregated to an enhanced global model with new training data.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Risk-based optimization of resource provisioning in mobile edge computing 3

Risk-based optimization of resource provisioning in mobile e...

引用

3rd acm/ieee symposium on edge computing, SEC 2018

作者： Badri, Hossein Bahreini, Tayebeh Grosu, Daniel Yang, Kai Department of Industrial and Systems Eng Wayne State University Detroit United States

ISBN: (纸本)9781538694459

In this paper, we propose a risked-based optimization approach to resource provisioning in MEC systems with the aim of taking into account the risk of overloading of edge servers when making allocation decisions. Assuming that resource requirements of mobile applications are stochastic parameters, we formulate the problem as a chance-constrained stochastic program. In order to solve the problem in reasonable amount of time, we employ the Sample Average Approximation (SAA) method. We evaluate the efficiency of the proposed approach by conducting an experimental analysis on instances with different problem settings. © 2018 ieee.

关键词： Mobile edge computing

来源：评论

学校读者我要写书评

暂无评论

Distributed Deep Neural Network Training on edge Devices 19

Distributed Deep Neural Network Training on Edge Devices

引用

4th acm/ieee symposium on edge computing (SEC)

作者： Benditkis, Daniel Keren, Aviv Mor-Yosef, Liron Avidor, Tomer Shoham, Neta Tal-Israel, Nadav Edgify Ramat Gan Israel

ISBN: (纸本)9781450367332

Deep Neural Network (Deep Learning) models have been traditionally trained on dedicated servers, after collecting data from various edge devices and sending them to the server. In recent years new methodologies have emerged for training models in a distributed manner over edge devices, keeping the data on the devices themselves. This allows for better data privacy and reduces the training costs. One of the main challenges for such methodologies is reducing the communication costs to and mainly from the edge devices. In this work we compare the two main methodologies used for distributed edge training: Federated Learning and Large Batch Training. For each of the methodologies we examine their convergence rates, communication costs, and final model performance. In addition, we present two techniques for compressing the communication between the edge devices, and examine their suitability for each one of the training methodologies.

关键词： edge device neural network deep learning federated learning large batch communication compression

来源：评论

学校读者我要写书评

暂无评论

Towards Scalable edge-Native Applications 19

Towards Scalable Edge-Native Applications

引用

4th acm/ieee symposium on edge computing (SEC)

作者： Wang, Junjue Feng, Ziqiang George, Shilpa Iyengar, Roger Pillai, Padmanabhan Satyanarayanan, Mahadev Carnegie Mellon Univ Pittsburgh PA 15213 USA Intel Labs Santa Clara CA USA

ISBN: (纸本)9781450367332

Latency-sensitive edge-native applications may be the key to commercial success of edge infrastructure. However, success in the form of widespread deployment of such applications poses its own challenges. These applications are edge-dependent by definition, and therefore cannot simply fail over to the cloud if the edge is overloaded. In this paper, we propose an adaptation-based strategy to allow scaling up the number of concurrent edge-native applications on a resource-limited cloudlet and wireless network. We demonstrate up to 40% reduction in offered load with minimal impact on latency on a variety of cognitive assistance tasks over non-adaptive approaches. Our approach is able to gracefully degrade and maintain quality of service for a subset of applications in the face of severely loaded conditions.

关键词： Wearable Cognitive Assistance Cloudlet edge computing Resource Management Mobile computing Gabriel

来源：评论

学校读者我要写书评

暂无评论

Zone of Control: The Basic computing Unit for Web of Everything 19

Zone of Control: The Basic Computing Unit for Web of Everyth...

引用

4th acm/ieee symposium on edge computing (SEC)

作者： Peng, Xiaohui Shi, Weisong Chinese Acad Sci Inst Comp Technol Beijing Peoples R China Wayne State Univ Detroit MI USA

ISBN: (纸本)9781450367332

In the era of Web of Everything (WoE), numerous things devices are connected to the Internet. They produce massive data and provide Web-based services at the network edge. New computing paradigms, such as edge computing, have been proposed to handle this data as close to where they are generated as possible. However, the Paradox of Classis Insecta (PCI) problem has become a hindrance of processing them on things devices located in the network edge. In this paper, we discuss the PCI problem, and propose Zone of Control (ZoC) architecture as a basic computing unit to address it. Unlike personal computer (PC) Web and mobile Web which are unified by their own hardware-software stack and REST architecture style, WoE consists of various domain-specific and physical-environment-oriented small distributed systems and is hard to form a unified technology stack. We need a basic system architecture abstraction to guide developers in the building of a more intelligent and efficient WoE system. The fundamental idea of ZoC is to divide WoE into multiple domain-specific zones with same or similar management, privacy and security policies. We introduce ZoC algebra with four operators to provide theoretical support for manipulating zones at system runtime. Early experiment results show that ZoC is beneficial to reducing overall data transmission and protecting privacy.

关键词： Web of Everything Internet of Things Zone of Control edge computing

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：