面向云平台的用户体验质量监测和管理技术研究
发布时间：2020-10-29 23:14
目前,云平台根据服务提供商和用户之间签订的服务水平协议(SLA)为用户提供付费服务,云平台提供商为了提供高质量的服务引入体验质量(QoE)指标,该指标包含用户关于服务质量(QoS)和用户需求的反馈信息。但有些贪婪的用户为了获取超出SLA以外的服务,故意提供错误的反馈信息,影响到QoE收集的准确性,因此,云平台提供商必须监控QoE指标,并根据该指标提供更好的服务。目前,已有研究者对云平台QoE的监测开展了大量研究工作,构建测量框架对用户的QoE进行有效监测与分析,但是还缺少综合的QoE测量与分析框架,以及从多层次、多粒度、宏观与微观结合的测量体系研究。为此,本文以为用户提供更好的QoS为目标。首先,提出基于代理的云平台体验质量监控框架(QoC),该框架从云端、传输网络与服务交互的客户端等多层次进行QoE相关技术参数测量,区分正面和负面的QoE,能够根据用户需求监控和管理云服务中的QoE,实现QoE预测和管理;然后,从用户角度测量不同终端设备在访问云服务时,RAM和虚拟内存等存储参数对QoE的影响,以及测量终端用户与云平台间网络距离以及不同视频、图像编码与压缩机制对QoE的影响;最后,根据QoE测量结果给出云平台管理建议。本文的详细研究工作如下:(1)鉴于前人提出的云平台QoE测量框架存在缺少因SLA违规而产生无效数据而无法修复和更新的功能,以及缺乏监视外部网络和客户端设备资源的性能,本文提出面向云平台的QoC框架来监控使用图像、视频流多媒体服务云用户的QoE,该框架提供在线运行策略更新,从内部自动监控云端到用户端的测量数据,结合用户提交的主观QoE与基于代理的测量框架测量的客观QoE数据,进行分析和管理,向云管理者或用户发送告警信息。因此,该QoC框架具有对客户端设备、应用服务等进行实时测量监控的功能,并实现在线更新服务。(2)为了进一步提升QoC框架对QoE测量的准确性,QoC框架采用多层次的体系结构,包括服务器端、代理端、缓存层。其中,代理端采用基于反向代理技术来构建具有负载均衡的服务器集群,保障多点测量时访问的时效性。该集群包含的缓存加速层,可减少后端数据库的负载。另外,构建流媒体数据库服务器集群层,该层应用主从同步机制,形成数据存储层,实现数据库读写分离和数据冗余。同时,提出一种适应视频文件大小的缓存替换算法,解决终端访问云流媒体服务时的性能瓶颈问题。因此,服务器端分层体系结构提高了整个系统的性能和稳定性,并具有高度的可扩展性,为将来扩展系统业务逻辑和增加用户量奠定了坚实的基础。(3)为了测量终端用户与云平台间网络距离对QoE的影响,本文测量了来自不同网络距离的云媒体视频的QoE。首先,测量3G/4G网络分别收集来自多个位置的云平台的性能指标。然后,基于不同云服务商、具有不同的编解码机制、比特率和分辨率的视频,进行主观QoE测量。结果表明,网络距离的远近直接影响到QoE水平。其次,测量用户的不同终端设备访问云服务时,RAM和虚拟内存等存储参数对用户体验产生的影响。结果表明,用户设备的闲置资源使用情况直接影响到云端在线流媒体播放效果,并用户的对访问速度和内容加载等方面影响较大。(4)为了评价在云平台中影响用户体验质量的因素,本文对社交云平台上图像/视频服务的QoE指标进行测量,并使用以节省存储和降低网络流为目的的压缩视频来提高服务的QoS。通过对影响因素进行分析,视频压缩降低了在线播放的质量,并影响下载之后的视频参数,从而降低了用户的QoE。为了进一步测量使用视频压缩后流媒体服务的QoE,通过上传、共享和播放社交云上不同的视频进行QoE指标评价。在上传图像期间,社交云平台会通过自动压缩原始图像来节省存储空间从而加速网页内容下载速度,但却降低了用户体QoE指标。通过实验测量不同云服务提供商图片压缩方法与视频流在不同分辨率下对用户体验质量的影响,给出客观评价,得出Tumblr、Twitter云端在线播放视频流质量较差,不能满足用户需求。
【学位单位】：哈尔滨工业大学
【学位级别】：博士
【学位年份】：2019
【中图分类】：F274;TP393.09;TP391.41
【文章目录】：
摘要
Abstract
List of Abbreviations and Glossary
Chapter1 Introduction
1.1 Problem Statement
1.2 Motivation
1.3 Aims and Objectives
1.4 Contribution of the thesis
1.5 Outline of Thesis
Chapter2 Background of QoE Framework for Multimedia in Cloud Computing
2.1 Introduction
2.2 Multimedia Quality Measurement
2.2.1 Subjective Multimedia Quality Measurement
2.2.2 Objective Multimedia Quality Measurement
2.3 Quality of Experience Frameworks for Multimedia Services
2.3.1 Quality of Service（QoS）
2.3.2 Quality of Experience（QoE）
2.4 Cloud Computing
2.4.1 Overview of the Cloud Infrastructure
2.4.2 Software as a Service（SaaS）
2.4.3 Platform as a Service（PaaS）
2.4.4 Infrastructure as a service（IaaS）
2.5 Summary
Chapter3 Qo C: Monitoring and Managing Quality of Experience （Qo E） Framework for Cloud Computing
3.1 Introduction
3.2 Related Works
3.3 Comparison of Qo E Framework with Qo C
3.4 Qo C framework Requirements
3.5 Design and Implementation of Qo C
3.6 Formal Description of Qo E Classic Architecture
3.7 Formal Description of Qo C Framework
3.8 Modeling and Analyzing Qo C Framework Overheads in Cloud environments
3.9 Mobile Application Experiments Based on Qo C
3.10 Summary
Chapter4 QoC Framework Implementation and Testing
4.1 Introduction
4.2 Problem Statement
4.3 Cloud Video Server Description
4.3.1 Web server
4.3.2 Database server
4.3.3 Content server
4.4 Related works
4.5 Hierarchical Design of the Server
4.5.1 Front-end access layer
4.5.2 Cache layer
4.5.3 Data storage layer
4.6 Implementation of Front-end access layer
4.7 Implementation of the cache layer
4.7.1 Acceleration Policy for Cache Layer
4.7.2 Caching Substitution Algorithm for Content Caching Layer
4.8 Implementation of Data Storage Layer
4.9 The Realization of Business Logic
4.10 System Performance Testing
4.10.1 Front End Access Layer Performance Test
4.10.2 Cache Layer Performance Test
4.10.3 Data Layer Performance Testing
4.10.4 Cache Replacement Algorithm Test
4.11 Summary
Chapter5 Monitoring Effect of Cloud Distance and Impact of Storage on End User’s QoE
5.1 Introduction
5.2 Problem Statement
5.3 Related Works
5.4 QoE Assessment from End User-side Measurement
5.4.1 QoE Assessment of Cloud Distance
5.4.2 QoE Assessment Design of Storage
5.5 QoE Evaluation Experiments Using Different Devices
5.5.1 QoE Experiment with HTC
5.5.2 QoE experiment with Samsung
5.6 Performance Experiments Results Analysis and Discussion
5.7 Summary
Chapter6 QoE Monitoring and Assessment of Video and Image Quality in Social Clouds
6.1 Introduction
6.2 Related Works
6.3 QoE Assessment Design of Video Quality
6.3.1 ACR Experiment Methodology
6.3.2 Selection of videos
6.4 QoE Assessment Experiments on Video Quality
6.4.1 Comparison of Cloud Compressed Videos Parameters
6.4.2 Evaluation of Mean Opinion Score
6.5 QoE Assessment Experiments on Image Quality
6.5.1 Samples Description and Comparison Experiments
6.5.2 Evaluation of User Ratings
6.6 Summary
Conclusion
Future work
Author’s Contribution
References
List of Publications
Under Review Papers
Acknowledgements
Resume
本文编号：2861605
【学位单位】：哈尔滨工业大学
【学位级别】：博士
【学位年份】：2019
【中图分类】：F274;TP393.09;TP391.41
【文章目录】：
摘要
Abstract
List of Abbreviations and Glossary
Chapter1 Introduction
1.1 Problem Statement
1.2 Motivation
1.3 Aims and Objectives
1.4 Contribution of the thesis
1.5 Outline of Thesis
Chapter2 Background of QoE Framework for Multimedia in Cloud Computing
2.1 Introduction
2.2 Multimedia Quality Measurement
2.2.1 Subjective Multimedia Quality Measurement
2.2.2 Objective Multimedia Quality Measurement
2.3 Quality of Experience Frameworks for Multimedia Services
2.3.1 Quality of Service（QoS）
2.3.2 Quality of Experience（QoE）
2.4 Cloud Computing
2.4.1 Overview of the Cloud Infrastructure
2.4.2 Software as a Service（SaaS）
2.4.3 Platform as a Service（PaaS）
2.4.4 Infrastructure as a service（IaaS）
2.5 Summary
Chapter3 Qo C: Monitoring and Managing Quality of Experience （Qo E） Framework for Cloud Computing
3.1 Introduction
3.2 Related Works
3.3 Comparison of Qo E Framework with Qo C
3.4 Qo C framework Requirements
3.5 Design and Implementation of Qo C
3.6 Formal Description of Qo E Classic Architecture
3.7 Formal Description of Qo C Framework
3.8 Modeling and Analyzing Qo C Framework Overheads in Cloud environments
3.9 Mobile Application Experiments Based on Qo C
3.10 Summary
Chapter4 QoC Framework Implementation and Testing
4.1 Introduction
4.2 Problem Statement
4.3 Cloud Video Server Description
4.3.1 Web server
4.3.2 Database server
4.3.3 Content server
4.4 Related works
4.5 Hierarchical Design of the Server
4.5.1 Front-end access layer
4.5.2 Cache layer
4.5.3 Data storage layer
4.6 Implementation of Front-end access layer
4.7 Implementation of the cache layer
4.7.1 Acceleration Policy for Cache Layer
4.7.2 Caching Substitution Algorithm for Content Caching Layer
4.8 Implementation of Data Storage Layer
4.9 The Realization of Business Logic
4.10 System Performance Testing
4.10.1 Front End Access Layer Performance Test
4.10.2 Cache Layer Performance Test
4.10.3 Data Layer Performance Testing
4.10.4 Cache Replacement Algorithm Test
4.11 Summary
Chapter5 Monitoring Effect of Cloud Distance and Impact of Storage on End User’s QoE
5.1 Introduction
5.2 Problem Statement
5.3 Related Works
5.4 QoE Assessment from End User-side Measurement
5.4.1 QoE Assessment of Cloud Distance
5.4.2 QoE Assessment Design of Storage
5.5 QoE Evaluation Experiments Using Different Devices
5.5.1 QoE Experiment with HTC
5.5.2 QoE experiment with Samsung
5.6 Performance Experiments Results Analysis and Discussion
5.7 Summary
Chapter6 QoE Monitoring and Assessment of Video and Image Quality in Social Clouds
6.1 Introduction
6.2 Related Works
6.3 QoE Assessment Design of Video Quality
6.3.1 ACR Experiment Methodology
6.3.2 Selection of videos
6.4 QoE Assessment Experiments on Video Quality
6.4.1 Comparison of Cloud Compressed Videos Parameters
6.4.2 Evaluation of Mean Opinion Score
6.5 QoE Assessment Experiments on Image Quality
6.5.1 Samples Description and Comparison Experiments
6.5.2 Evaluation of User Ratings
6.6 Summary
Conclusion
Future work
Author’s Contribution
References
List of Publications
Under Review Papers
Acknowledgements
Resume
本文编号：2861605
本文链接：https://www.wllwen.com/guanlilunwen/glzh/2861605.html
教材专著
