A feedback control model for resource management on streaming media servers

Utilizing system resource efficiently, and guaranteeing the client's basic QoS requirement, and supporting more requirements concurrently, is very important to streaming media applications. In this paper, we present a feedback control resource management model and resource controllers for streaming servers. Through sampling the system states periodically, and adjusting the service attributes to change the resource utilization in the next sampling period, this model guarantees the basic QoS requirement, improves the resource utilization and concurrency. We analysis the model and evaluate its performance.