Bandwidth constrained smoothing for multimedia streaming with scheduling support

Providing a satisfactory multimedia service in networking environments requires an effective media delivery mechanism. However, a common network such as the Internet does not provide a guaranteed network bandwidth to accommodate multimedia service in a reliable fashion. A typical approach to assist multimedia delivery is via buffer management and task scheduling in end-systems. Buffer management techniques are classified into two categories; one is to adapt the changes in network load and the other is to smooth the bandwidth requirement. The former may cause a serious loss of service quality whereas the latter is unable to adapt to the dynamic network condition. In this paper, we propose a bandwidth-adaptive media smoothing technique which smoothes the bandwidth requirement for media delivery at run time by considering the availability of network bandwidth. Meanwhile, the bandwidth smoothing technique still has the possibility of causing jitter because the policy runs on the application layer so that it cannot guarantee task completion in time. Thus, we also propose a task scheduling algorithm optimized for the bandwidth adaptive smoothing. This scheduling technique handles the media data appropriately in order to minimize jitter. Simulation results with prerecorded MPEG videos show that the quality of delivered video is improved with the proposed bandwidth adaptive smoothing and task scheduling mechanisms.