An SDN-based device-aware live video service for inter-domain adaptive bitrate streaming
Zahran, Ahmed H.
Sreenan, Cormac J.
Association for Computing Machinery (ACM)
The emerging popularity of live streaming services poses a great challenge for the rigid and static traditional Internet architecture. The rise in adaptation of Software Defined Networking (SDN) by Internet Service Providers (ISP) and Content Delivery Networks (CDN) presents an opportunity to dynamically adapt and respond in real-time to high definition (HD) mega events or dynamic short-lived broadcast events. In this paper, we present an SDN-based system design that utilizes a communication framework between ISPs and CDNs to interact and thus enable a reliable and resource efficient live streaming service. We build and deploy an optimization model that can maximize the video quality for users while minimizing the resource utilization for both ISPs and CDNs. The model considers device capabilities, network constraints and the subscription level of users with the ISP/CDN. Our system is a network-assisted, cross-layer, approach that implements multicast at the network layer and can dynamically adapt the video bitrates that are served to each client at the application layer. We build a prototype of our proposed design and evaluate real-world scenarios with up to 500 users streaming multiple videos at different bitrates. Results show that our approach can increase average user goodput by up to 70% while almost eliminating frame drops by handling network congestion.
Network-assisted solutions , ISP and CDN collaboration , Software defined networking , Live video , Congestion handling , Network layer multicast , Optimization models , Video streaming
Khalid, A., Zahran, A. H. and Sreenan, C. J. (2019) 'An SDN-based device-aware live video service for inter-domain adaptive bitrate streaming', Proceedings of the 10th ACM Multimedia Systems Conference, Amherst, Massachusetts, 18-21 June, pp. 121-132. doi: 10.1145/3304109.3306229
© 2019, Association for Computing Machinery. This is the author's version of the work. It is posted here for your personal use. Not for redistribution. The definitive Version of Record was published in Proceedings of the 10th ACM Multimedia Systems Conference: https://doi.org/10.1145/3304109.3306229