Analyzing TCP Throughput Stability and Predictability with Implications for Adaptive Video Streaming

TL;DR

This paper investigates TCP throughput stability and predictability in video sessions, proposing a hidden Markov model-based prediction method that enhances adaptive video streaming quality.

Contribution

It provides the first large-scale measurement analysis of throughput stability and introduces a novel prediction mechanism for better video bitrate adaptation.

Findings

Throughput stability varies significantly across sessions

The hidden Markov model outperforms existing prediction methods

Improved predictions lead to better user experience in streaming

Abstract

Recent work suggests that TCP throughput stability and predictability within a video viewing session can inform the design of better video bitrate adaptation algorithms. Despite a rich tradition of Internet measurement, however, our understanding of throughput stability and predictability is quite limited. To bridge this gap, we present a measurement study of throughput stability using a large-scale dataset from a video service provider. Drawing on this analysis, we propose a simple-but-effective prediction mechanism based on a hidden Markov model and demonstrate that it outperforms other approaches. We also show the practical implications in improving the user experience of adaptive video streaming.