A Novel Dynamically Optimized Embedded Video Burst Scheduler to Enhance the System QoS
Real-time video processing is still now a formidable task for the strict requirement on latency control and packet loss minimization. Burst processing has come to the rescue by offering buffer less operation and separation of control and data information. In this paper a novel dynamically-optimized embedded burst scheduling method suitable for processing class-differentiated video channels has been proposed. The method is based on statistical Markov chains where the initial scheduled Markov transition probabilities are subsequently adaptively reconfigured by the central scheduler to maintain the best system Quality of Service (QoS).