算力网络支撑下的泛在化视频传输调度  

Ubiquitous Video Transmission Scheduling Supported by Computing Power Network

作　　者：张旭光 陈鸣锴[1,2] 魏昕 Zhang Xuguang;Chen Mingkai;Wei Xin(College of Telecommunications&Information Engineering,Nanjing University of Posts and Telecommunications,Nanjing 210003;Key Laboratory of Broadband Wireless Communication and Sensor Network Technology(Nanjing University of Posts and Telecommunications),Ministry of Education,Nanjing 210003)

机构地区：[1]南京邮电大学通信与信息工程学院,南京210003 [2]宽带无线通信与传感网技术教育部重点实验室(南京邮电大学),南京210003

出　　处：《计算机研究与发展》2023年第4期786-796,共11页Journal of Computer Research and Development

基　　金：国家自然科学基金项目(62071254,62231017,62001246);江苏高校优势学科建设工程项目;南京邮电大学引进人才自然科学研究启动基金(NY222035);中国博士后科学基金资助项目(2022M721694);江苏省卓越博士后计划(2022ZB398)。

摘　　要：视频数据量的爆炸式增长、视频形式愈加多样、视频业务的泛在化是当前视频通信技术发展的三大特点,这无疑会导致核心网过载和视频传输调度难的问题.为了缓解这些问题,提出一种算力网络支撑下的泛在化视频传输调度方案.具体地,首先提出一种视频层次化编解码模型提升视频内容部署和传输的灵活性;其次,考虑视频业务的差异化指标约束,提出面向业务的有效吞吐量模型;最后,借助算力网络的支撑,一方面通过任务分解来有效利用“碎片化”的网络资源,另一方面通过对网络状态的全局检测和实时感知,实现视频内容的精准部署和网络资源的高效调配.仿真实验验证了所提方案在核心网流量卸载、有效吞吐量提升、网络资源利用率提升方面的有效性.The explosive growth of video data volume,the increasing diversity of video forms,and the ubiquity of video services are the three main characteristics of the development of current video communication technology.This fact will undoubtedly lead to two main problems.The first is that the traffic burden of the core network is difficult to offload,and the second is that,the video transmission conflict is difficult to coordinate.In order to alleviate these two problems,we propose a ubiquitous video scheduling scheme,supported by computing power networks.Specifically,we first propose a hierarchical video coding and decoding model to enhance the flexibility of video content deployment and transmission;Secondly,we propose a service-oriented good-put model with the consideration that the diversity transmission parameter constraints of different video services;Finally,with the support of computing power network,on the one hand,through task decomposition,the“fragmented”network resources can be utilized effectively,and on the other hand,through global detection and real-time perception of network status,accurate deployment of video content and efficient scheduling of network resources can be achieved.Simulation experimental results verify the effectiveness of the proposed scheme in terms of core network traffic offloading,good-put improvement,and network resource utilization rate improvement.

关 键 词：算力网络 泛在化视频 视频调度 有效吞吐量 内容分发 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]