Vortragender | Dr. Mu Mu |
Titel | Fairness delivered: SDN-assisted network resource allocation for adaptive video streaming |
Zeit | Mittwoch, 18.02.2015, 14:00 – 16:00 Uhr |
Ort |
S3|20 / Raum 111 Rundeturmstraße 10, 64283 Darmstadt |
Abstract |
|
MPEG-DASH as an HTTP adaptive streaming (HAS) technology is becoming a popular vehicle for online video delivery, thanks to its unique adaptation feature that allows dynamic selection of quality representations through bandwidth estimation. Recent years have seen a number of single-stream HAS optimization algorithms emerging, whose main objective is to maximize the quality of user experience on a single HAS stream with respect to the dynamics of network resources. However such optimization algorithms work independently (and sometimes aggressively) without any coordination with other devices in the same network. This leads to QoE fluctuations and unfairness when multiple HAS streams compete for constraint network resources. Recent advancements in networking such as the software-defined networking (SDN) and network-function virtualization (NFV) enable network-wide flexibility and programmability allowing comprehensive network and service functions to be deployed easily in an on-demand fashion. Leveraging the aforementioned technologies, we designed a QoE network resource allocation framework, which is encapsulated in the model of QoE-as-a-service. Employing a number of tailored QoE metrics (such as video quality, switching impact, and cost efficiency), the framework orchestrates network resource provisioning between HAS media streams with respect to user perception of video quality in order to maximize user-level QoE fairness and to reduce unnecessary network fluctuations. In this talk, I'll introduce the design and implementation of the framework accompanied with initial experimental results. |
|
Bio | |
I am a Senior Research Associate in the School of Computing and Communications at Lancaster University, UK where I also completed PhD of Computer Science on the topic of QoE in content distribution networks. Prior to joining Lancaster University, I studied in Germany and my Master’s degree of Information and Communication Engineering was awarded by TU-Darmstadt (Darmstadt University of Technology) following an internship at Deutsche Telekom Technology Centre. I have been actively contributing to proposal writing, research, and project management of multiple European and UK projects such as MediaCityUK FIRM, EC FP7 P2P-Next, FP7 STEER, FP7 OFELIA, and most recently FP7 FI-CONTENT 2. I am also leading the R&D of the Lancaster University Vision Social IPTV system and Storisphere storytelling platform. My current research interests cover the topics of QoE-aware content distribution, SDN/NFV applications, immersive experience through media orchestration, creative media and storytelling as well as improve user engagement by understanding the patterns of web browsing and online TV viewing. I published over 40 research papers in international conferences and journals of high impacts, including IEEE Transactions on Multimedia, IEEE Communications Magazine, IEEE Multimedia Magazine, ACM Conference on Multimedia and ACM Multimedia Systems. I regularly serve as programme committee member and reviewer of workshops, conferences and journals such as Springer MMSJ, IEEE TMM, ACM MMSYS, ACM TVX, IEEE ICC, and Elsevier COMNET. I also provide reviews for book proposals and project proposals. |
|