Real-time scheduling based on imprecise computation for scalable streaming media system over the Internet

This paper proposes a performance metrics and a real-time scheduling algorithm based on imprecise computation workload model for delivery of scalable streaming media, which can be adapted to network status and QoS requirement over the best-effort Internet. The scheduling task of a scalable streaming media is partitioned into two subtasks: the mandatory subtask for the base layer and the optional subtask for the enhancement layers. The imprecise computation workload model and real-time scheduling algorithm provide scheduling flexibility by trading off video quality reconstructed in client to meet the playback deadline. Thus, the better usage of available bandwidth and smoother playback are achieved.