A Multi-Factor QoE Model for Adaptive Streaming over Mobile Networks

In the context of HTTP Adaptive Streaming (HAS), a main challenge is how to evaluate the quality of experience (QoE) of a session. In general, the QoE of a streaming session is affected by three key factors, namely initial delay, varying perceptual quality, and interruptions. In this paper, in order to investigate the impacts of quality variations and interruptions on the QoE of a session, we propose to use histograms, which represent the frequency distributions of these recurrent factors in a session. The QoE of a session is modeled by pooling the statistics of the recurrent factors and the initial delay. Evaluation results show that the proposed model has a high prediction performance. Especially, the proposed model provides insights into the impacts of different influence factors, thus leading to suggestions to improve the quality of streaming video.