摘要
由于疫情原因,基于WebRTC技术的视频会议得到快速发展,给人们的日常生活带来极大的便利.但是视频会议规模受限于客户端的网络状况以及服务器的上行带宽,对于中小企业开发视频会议系统是一个挑战.因此,为了改善视频会议整体性能,本文在Simulcast+SFU架构的多人视频会议基础上,同时考虑客户端带宽消耗比、视频质量、码率差和服务器上行带宽消耗建立多码率优化高维多目标优化模型.另外,在对模型求解中,本文提出了一种基于适应度,成绩标量函数(ASF),角度信息以及收敛偏好函数(FC)的多指标交配池选择高维多目标优化算法(MAOEA-MIPS).仿真结果表明,模型在保证视频质量的同时,有效地减少了客户端带宽消耗以及服务器上行带宽消耗.通过与其他高维多目标优化算法对比,表明MAOEA-MIPS在求解模型中可以获得更好的结果。
Due to the pandemic,WebRTC technology-based video conference developed rapidly and brought great convenience to peo-ples daily lives.However,the scale of video conference is limited by the client network and the upstream bandwidth of the server,which is a challenge for small and medium-sized enterprises to develop video conference systems.Therefore,to improve the overall performance of video conference,this paper is based on the Simulcast+SFU architecture of multi-party video conference,and establi-shes a multi-bitrate optimization many-objective model which simultaneously considers four target models:client bandwidth consump-tion ratio,video quality,bitrate difference,and server uplink bandwidth consumption.To solve the model,this paper proposes a multi-index mating pool selection for many-objective optimization algorithm(MAOEA-MIPS)based on fitness,achievement scalarizing func-tion(ASF),angle information and favorable convergence(FC).The simulation results indicate that the model effectively reduces the bandwidth consumption of the client and the upstream bandwidth of the server while ensuring the quality of the video.Compared with other many-objective optimization algorithms,it shows that MAOEA-MIPS can obtain better results in solving the model.
作者
高朝龙
张景波
蔡星娟
GAO Chao-long;ZHANG Jing-bo;CAI Xing-juan(College of Computer Science and Technology,Taiyuan University of Technology,Taiyuan 030024,China)
出处
《小型微型计算机系统》
CSCD
北大核心
2023年第8期1735-1742,共8页
Journal of Chinese Computer Systems
基金
国家自然科学基金青年科学基金项目(61806138)资助
山西省重点研发计划项目(201903D421003)资助。
