学术文献库

HTTP Adaptive Streaming (HAS) is becoming the de-facto video delivery technology over best-effort networks nowadays, thanks to the myriad advantages it brings. However, many studies have shown that HAS suffers from many Quality of Experience (QoE)-related issues in the presence of competing players. This is mainly caused by the selfishness of the players resulting from the decentralized intelligence given to the player. Another limitation is the bottleneck link that could happen at any time during the streaming session and anywhere in the network. These issues may result in wobbling bandwidth perception by the players and could lead to missing the deadline for chunk downloads, which result in the most annoying issue consisting of rebuffering events. In this paper, we leverage the SoftwareDefined Networking paradigm to take advantage of the global view of the network and its powerful intelligence that allows reacting to the network changing conditions. Ultimately, we aim at preventing the re-buffering events, resulting from deadline misses, and ensuring high QoE for the accepted clients in the system. To this end, we use Deterministic Network Calculus (DNC) to guarantee a maximum delay for the download of the video chunks while maximizing the perceived video quality. Simulation results show that the proposed solution ensures high efficiency for the accepted clients without any rebuffering events which result in high user QoE. Consequently, it might be highly useful for scenarios where video chunks should be strictly downloaded on-time or ensuring low delay with high user QoE such as serving video premium subscribers or remote control/driving of an autonomous vehicle in future 5G mobile networks.