Scalable video coding(SVC)has been widely used in video-on-demand(VOD)service,to efficiently satisfy users’different video quality requirements and dynamically adjust video stream to timevariant wireless channels.Und...Scalable video coding(SVC)has been widely used in video-on-demand(VOD)service,to efficiently satisfy users’different video quality requirements and dynamically adjust video stream to timevariant wireless channels.Under the 5G network structure,we consider a cooperative caching scheme inside each cluster with SVC to economically utilize the limited caching storage.A novel multi-agent deep reinforcement learning(MADRL)framework is proposed to jointly optimize the video access delay and users’satisfaction,where an aggregation node is introduced helping individual agents to achieve global observations and overall system rewards.Moreover,to cope with the large action space caused by the large number of videos and users,a dimension decomposition method is embedded into the neural network in each agent,which greatly reduce the computational complexity and memory cost of the reinforcement learning.Experimental results show that:1)the proposed value-decomposed dimensional network(VDDN)algorithm achieves an obvious performance gain versus the traditional MADRL;2)the proposed VDDN algorithm can handle an extremely large action space and quickly converge with a low computational complexity.展开更多
信息中心网络(Information Centric Network,ICN)因其内容分发的高效性,被认为是下一代互联网的原型之一。因此,重点研究ICN网络中可伸缩视频(Scalable Video Coding,SVC)分发中不同缓存策略下的用户体验质量(Quality of Experience,Qo...信息中心网络(Information Centric Network,ICN)因其内容分发的高效性,被认为是下一代互联网的原型之一。因此,重点研究ICN网络中可伸缩视频(Scalable Video Coding,SVC)分发中不同缓存策略下的用户体验质量(Quality of Experience,Qo E)与成本。一方面,根据ICN网络和SVC视频缓存的特点,设计了一个ICN分层视频分发系统,并建立数学模型对用户体验质量和路由成本进行刻画;另一方面,通过在ndn SIM仿真平台,对不同缓存大小下的Fifo、Lfu、Lru三种替换策略进行比较,结果显示,随着缓存的不断增大,Qo E整体呈现上升趋势,成本整体呈现下降趋势。展开更多
The exponential growths of the World Wide Web (WWW) users have made the deployment of proxy servers popular on a network with limited resources. WWW clients perceive better response time, improved performance and spee...The exponential growths of the World Wide Web (WWW) users have made the deployment of proxy servers popular on a network with limited resources. WWW clients perceive better response time, improved performance and speed when response to requested pages are served from the cache of a proxy server, resulting in faster response times after the first document fetch. This work proposes cyclic multicast as a scalable technique for improving proxy server performance for next generation networks. The proposed system uses a cyclic multicast engine for the delivery of popular web pages from the proxy server cache to increasingly large users under limited server capacity and network resources. The cyclic multicast technique would be more efficient for the delivery of highly requested web pages from the cache to large number of receivers. We describe the operation of the cyclic multicast proxy server and characterized the gains in performance.展开更多
基金supported by the National Natural Science Foundation of China under Grant No.61801119。
文摘Scalable video coding(SVC)has been widely used in video-on-demand(VOD)service,to efficiently satisfy users’different video quality requirements and dynamically adjust video stream to timevariant wireless channels.Under the 5G network structure,we consider a cooperative caching scheme inside each cluster with SVC to economically utilize the limited caching storage.A novel multi-agent deep reinforcement learning(MADRL)framework is proposed to jointly optimize the video access delay and users’satisfaction,where an aggregation node is introduced helping individual agents to achieve global observations and overall system rewards.Moreover,to cope with the large action space caused by the large number of videos and users,a dimension decomposition method is embedded into the neural network in each agent,which greatly reduce the computational complexity and memory cost of the reinforcement learning.Experimental results show that:1)the proposed value-decomposed dimensional network(VDDN)algorithm achieves an obvious performance gain versus the traditional MADRL;2)the proposed VDDN algorithm can handle an extremely large action space and quickly converge with a low computational complexity.
文摘信息中心网络(Information Centric Network,ICN)因其内容分发的高效性,被认为是下一代互联网的原型之一。因此,重点研究ICN网络中可伸缩视频(Scalable Video Coding,SVC)分发中不同缓存策略下的用户体验质量(Quality of Experience,Qo E)与成本。一方面,根据ICN网络和SVC视频缓存的特点,设计了一个ICN分层视频分发系统,并建立数学模型对用户体验质量和路由成本进行刻画;另一方面,通过在ndn SIM仿真平台,对不同缓存大小下的Fifo、Lfu、Lru三种替换策略进行比较,结果显示,随着缓存的不断增大,Qo E整体呈现上升趋势,成本整体呈现下降趋势。
文摘The exponential growths of the World Wide Web (WWW) users have made the deployment of proxy servers popular on a network with limited resources. WWW clients perceive better response time, improved performance and speed when response to requested pages are served from the cache of a proxy server, resulting in faster response times after the first document fetch. This work proposes cyclic multicast as a scalable technique for improving proxy server performance for next generation networks. The proposed system uses a cyclic multicast engine for the delivery of popular web pages from the proxy server cache to increasingly large users under limited server capacity and network resources. The cyclic multicast technique would be more efficient for the delivery of highly requested web pages from the cache to large number of receivers. We describe the operation of the cyclic multicast proxy server and characterized the gains in performance.
