边缘计算中面向互动直播的用户分配策略  

User Allocation Strategy for Interactive Live Streaming in Edge Computing

作　　者：刘伟[1,2,3,4] 张骁宇[1] 杜薇[1,2] 彭若涛 Liu Wei;Zhang Xiaoyu;Du Wei;Peng Ruotao(School of Computer and Artificial Intelligence,Wuhan University of Technology,Wuhan 430070;Hubei Key Laboratory of Transportation Internet of Things(Wuhan University of Technology),Wuhan 430070;Key Laboratory of Embedded System and Service Computing(Tongji University),Ministry of Education,Shanghai 201804;State Key Lab of Processors(Institute of Computing Technology,Chinese Academy of Sciences),Beijing 100190)

机构地区：[1]武汉理工大学计算机与人工智能学院,武汉430070 [2]交通物联网湖北重点实验室(武汉理工大学),武汉430070 [3]嵌入式系统与服务计算教育部重点实验室(同济大学),上海201804 [4]处理器芯片全国重点实验室(中国科学院计算技术研究所),北京100190

出　　处：《计算机研究与发展》2023年第8期1858-1874,共17页Journal of Computer Research and Development

基　　金：湖北省自然科学基金面上项目(2020CFB749);同济大学嵌入式系统与服务计算教育部重点实验室开放基金项目(ESSCKF2018-05);中科院计算技术研究所计算机体系结构国家重点实验室开放课题(CARCHB202015)。

摘　　要：将互动直播部署在边缘计算环境中,可以在网络边缘对直播视频进行转码和传输,通过用户附近的边缘服务器提供低延迟的直播服务.然而,在多边缘服务器、多用户场景下存在着直播用户分配问题,导致直播用户体验质量(quality of experience, QoE)无法得到保证.为了提高直播用户QoE,需要根据用户的个性化需求合理地分配服务器资源.首先分析真实数据集,发现大多数用户处于多基站重叠覆盖区域内,并且不同用户的互动需求存在差异;然后根据互动直播的特点提出一种适用于边缘计算场景的用户QoE模型,该模型综合考虑了直播用户的视频质量和互动体验;最后设计一种高效的直播用户分配算法,优化了多边缘服务器重叠覆盖区域内的直播用户QoE.仿真实验表明,所提出的用户分配策略可为用户提供高码率和低延迟的直播视频,同时能有效降低边缘服务器切换次数和码率抖动,使直播用户QoE相较于其他策略提升超过19%.Deploying interactive live streaming in edge computing environment enable us to offload the transcoding and delivery cost to network edge and provide service with lower latency via the edge servers near the users.However,there is a problem of user allocation in the real complex multi-server and multi-user scenarios,leading to a poor quality of experience(QoE).In order to improve QoE of overlapping coverage areas,it is necessary to select edge servers for individual live streaming users according to their needs and to allocate server resources reasonably.First,the analysis of real-world data sets reveals that most users are in the overlapping coverage area of multiple base stations whose interaction needs vary.Then,a QoE model suitable for edge computing scenarios is proposed according to the analysis,which is based on the characteristics of interactive live streaming and comprehensively considers the interactive and video viewing experience of users.Finally,an efficient live streaming user allocation algorithm is designed to optimize the user QoE in the overlapping coverage area of multiple edge servers.Simulation experiments show that this strategy can provide users with high bit rate and low latency streaming while controlling the edge servers switching and bit rate jitter,thus improving the QoE of users by more than 19%compared with the other strategies.

关 键 词：互动直播 边缘计算 用户体验 重叠覆盖区域 用户分配 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]