QoS control for real-time VBR video traffic using prediction-based dynamic resource allocation architecture

In this paper, we propose an efficient quality providing scheme to satisfy a diversify combination of delay bound and loss ratio requirements from users by a predictive dynamic resource allocation method for real-time video applications. For utilizing the network resources more efficiently while meeting the service requirements, we adjust the resources based on the predicted traffic and the currently provided quality level. We developed a simple delayed-packet counter updating scheme for real time QoS monitoring. Simulation results show that our proposed method can provide an accurate and flexible quality control.