Video QoE Killer and Performance Statistics in WebRTC-based Video Communication

Abstract

Abstract: In this paper, we investigate session-related performance statistics of a Web-based Real-Time Communication (WebRTC) application called appear.in. We explore the characteristics of these statistics and explore how they may relate to users' Quality of Experience (QoE). More concretely, we have run a series of tests involving two parties and according to different test scenarios, and collected real-time session statistics by means of Google Chrome's WebRTC-internals tool. Despite the fact that the Chrome statistics have a number of limitations, our observations indicate that they are useful for QoE research when these limitations are known and carefully handled when performing post-processing analysis. The results from our initial tests show that a combination of performance indicators measured at the sender's and receiver's end may help to identify severe video freezes (being an important QoE killer) in the context of WebRTC-based video communication. In this paper the performance indicators used are significant drops in data rate, non-zero packet loss ratios, non-zero PLI values, and non-zero bucket delay.