2
Abstract:
Abstract--- Multimedia streaming applications consume a significant amount of server and network resources due to the high bandwidth and long duration of audio and video clips. Making streaming services economically viable requires techniques for minimizing the incremental cost of serving a new client, particularly for popular content. Patching [1] reduces server and network overhead by allowing a client to receive (part of) a multimedia stream by listening to an ongoing transmission of the same clip, without increasing client playback delay. However, existing patching schemes [1--3] do not fully exploit the client buffer space or the ability to listen to more than one ongoing transmission, for reducing bandwidth overheads. In this paper, we first introduce Periodic Buffer Reuse (PBR) patching that maximizes the amount of data that a client can retrieve from the ongoing transmission. Similar to the existing schemes, PBR can employ a threshold to determine when to start a new complete transmission of the stream. We derive a closed-form expression for the transmission bandwidth requirements for PBR patching, and show how to determine the optimal threshold value. Our performance comparison demonstrates that PBR can significantly outperform existing patching schemes. We then present Greedy Buffer Reuse (GBR), an algorithm that allows clients to patch to multiple ongoing transmissions. We show that this algorithm provably minimizes the server and network transmission bandwidth requirements. Simulation experiments demonstrate that GBR patching offers a sizeable reduction in transmission overhead over any of the threshold-based schemes, and rarely requires the client to listen to more than three simultaneous transmissions, for the scenarios we examine. I.
