Self-diagnostic peer-assisted video streaming through a learning framework

Quality control and resource optimization are challenging problems in peer-assisted video streaming systems, due to their large scales and unreliable peer behavior. Such systems are also prone to performance degradation in the event of drastic demand changes, such as ash crowds and large-scale simultaneous peer departures. In this paper, we demonstrate the de ciency of state-of-the-art video streaming systems by analyzing real-world traces from UUSee, a popular commercial P P media streaming system based in China, during the Beijing Olympics. We show how simple machine learning techniques combined with periodic collection of statistics can be used for automated monitoring and diagnosis of peer-assisted video streaming systems. With such a framework, it is possible to estimate performance given certain resource usage patterns, making resource utilization more e cient. It also enables the prediction of large-scale performance degradation due to irregular demand patterns. e e ectiveness ...

• Type: Conference
• Year: 2010
• Where: MM
• Authors: Di Niu, Baochun Li, Shuqiao Zhao
• Keywords: Large-scale Simultaneous Peer, MM 2010, Multimedia, Performance Degradation, Video Streaming Systems