The large amount of bandwidth and other resources required to deliver streaming media limits the number of concurrent users. We propose a virtual active network (VAN) architecture for streaming media data delivery over wide area networks. In the proposed architecture, cooperating proxies support multiplexing and delivery of live streaming media. The hierarchical delivery structure is dynamically adjusted based on user population distribution, usage patterns, and network conditions. The proposed system architecture provides (1) reliable and high quality live streaming media delivery; (2) lower server resource requirements at the content provider sites; (3) reduced inter-ISP traffic; (4) application level routing for rapid deployment; and (5) cost-effective media data delivery. To deal with one characteristics of live broadcasting events, burst traffic at the beginning of the events, our system features a unique function that dynamically clusters multiple proxy servers to form a server farm to handle a large number of user login events when needed. Experimental results show that the proposed VAN architecture consistently delivers reliable live data streams using resources within 10 percent of the theoretically possible lower bound. The experimental results also show the effectiveness of our load balance algorithms to handle various user patterns, server capacity, and network congestion events.
Optimal Proxy Cache Allocation for Efficient Streaming Media Distribution
