Performance Characterization of Low-Latency Adaptive Streaming From Video Portals

Abstract

News-based websites and portals provide significant amounts of multimedia content to accompany news stories and articles. In this context, the HTTP adaptive streaming is generally used to deliver video over the best-effort Internet, allowing smooth video playback and an acceptable Quality Of Experience (QoE). To stimulate the user engagement with the provided content, such as browsing between videos, reducing the videos’ startup time has become more and more important: while the current median load time is in the order of seconds, research has shown that the user waiting times must remain below two seconds to achieve an acceptable QoE. In this paper, four complementary components are optimized and integrated into a comprehensive framework for low-latency delivery of news-related video content: 1) server-side encoding with short video segments; 2) HTTP/2 server push at the application layer; 3) server-side user profiling to identify relevant content for a given user; and 4) client-side storage to hold proactively delivered content. Using a large data set of a major Belgian news provider, containing millions of text- and video-based article requests, we show that the proposed framework reduces the videos’ startup time in different mobile network scenarios by over 50%, thereby improving the user interaction and skimming available content.