摘要
网络多媒体内容的碎片性、动态内容不可缓存性等特性对传统内容分发提出了极大挑战。基于用户网页内容访问模式,本文采用WebRTC技术,利用浏览器协助机制,提出基于请求转发的碎片化动态内容分发算法。本文基于实际系统日志,研究用户的网页浏览行为(包括在线时长分布,重复访问规律)、应用层网页访问故障分布以及网络层网络状态变化,设计节点状态感知的转发节点选择算法。该算法基于网络层信息,同时考虑网页分发的新特性,提出请求转发机制。通过重构用户–用户–服务器的分发路径,恢复网页内容分发故障,“透明”增强用户端感知的系统可靠性。
The characteristics including the fragmentation and non-cacheable dynamic content of network multimedia present a great challenge to traditional content distribution. Based user's web browsing patterns, in this paper, we propose a dynamic content distribution algorithm based on request forwarding, with WebRTC technology and browser-assist mechanism. We analyse user’s browsing behaviors (e.g., including online duration and visiting frequency), application-layer web page request failure, network layer network condition change based on the real-world log system to design a condition-aware redirecting node selection algorithm. The algorithm utilizes the network-layer information, while considering the new characteristics of the Web, and proposes a request forwarding mechanism. By reconstructing the user-user-server distribution path and restoring the failure of web content distribution, "transparency"enhances the system reliability perceived by the client.
出处
《计算机科学与应用》
2020年第1期81-97,共17页
Computer Science and Application
