Our work on video delivery performance characterization and optimization has been accepted to appear in IEEE Trans. on Multimedia


Video delivery performance is the main factor that affects the Internet video quality. Characterizing the video delivery performance, especially the delivery throughput, can help content providers as well as Internet service providers (ISPs) in system optimization and network planning. This paper presents a comprehensive study of video delivery throughput of a large-scale commercial VoD (Video-on-Demand) system. The analysis is based on a unique dataset consisting of more than 20 million video download speed measurements from massive users over two weeks. We find that user speed exhibits a large variation over time of day as well as across provincial locations. In particular, the worst performance of day is 30\% lower than the peak performance. We also observe that video download speed has a notable influence on video quality and user engagement. The impact however becomes limited when the speed increases beyond certain threshold, which is mostly dependent on the video encoded bitrates. We further examine the interaction between Internet infrastructure and video delivery performance using linear regression model and find that crossing the ISP or regional network border yields 15-20\% speed loss. Based on these observations, we finally evaluate the potential of edge caching and hybrid CDN-P2P in the improvement of video download performance and video quality.


Zhenyu Li, Qinghua Wu, Kave Salamatian, and Gaogang Xie. Video Delivery Performance of a large-scale VoD system and the Implications on Content Delivery. To appear in IEEE Transactions on Multimedia, 2015 (camera-ready version will be soon available)