An Adaptive Control Scheme for Deadline-aware Streaming

Abstract

With increasing popularity on deadline-aware streaming service, how to deliver these multimedia data quickly and efficiently has become critical. Traditional transport protocols and congestion control schemes fail to meet expectations. On one hand, the traditional transport protocols like TCP are far away from minimizing the queuing delay. On the other hand, existing congestion control schemes is limited in transport layer, resulting in the neglect of features and deadline requirements in application layer. To address these issues, we propose a Heuristic Delay-based Adaptive Scheme, namely HDAS, based on Deadline-aware Transport Protocol (DTP). HDAS not only adjusts the window size and sending rate according to the network feedback but also schedules packets to improve the overall Quality of Experience (QoE). The proposed scheme adopts delay-based adaptive adjustment for congestion control and greedy strategy for scheduler. Extensive simulations demonstrate that our scheme outperforms most of the existing schemes on overall QoE.