Supporting Non-Linear and Non-Continuous Media Access in Peer-to-Peer Multimedia Systems

Abstract

With the proliferation of interactive media such as video-on-demand, massive multiplayer virtual environments, massively multiplayer online game, zoomable video, Google Earth, and free-viewpoint video, non-linear and non-continuous media accesses have become the norm. The peer-to-peer architecture is often adopted for scalable delivery of media content to potentially millions of users. Non-linear and non-continuous access patterns caused by massive user interactions, however, pose new challenges on peer-to-peer streaming systems. In this thesis, we first analytically study the effect of non-linear and non-continuous media accesses on the streaming system performance. Then, a novel user access pattern driven content discovery mechanism is proposed to reduce the content discovery time during non-continuous accesses. Finally, we devise a joint request and service scheduling scheme for non-linear access patterns. Our work in this thesis can help optimize peer-to-peer streaming systems towards better supporting non-linear and non-continuous access patterns, leading to improvement in user experience.