A Pre-buffer Mechanism in Multi-channel P2P Live Streaming

Abstract

In this paper, we mainly concern about the channel switch delay of multi-channel P2P live streaming system. After study the essential aspects of multi-channel video systems, including channel switching, peer churn, and channel popularity, we develop a new user behavior oriented and content oriented pre-buffer mechanism to improve the user experience in live streaming borrowing the idea of user behavior targeting. This new mechanism uses the user behavior records to make a prediction of user's viewing. It can get a pre-buffer succeed ratio about 40% on average, and make the channel switch delay much lower.