A Novel Strategy to Evaluate QoE for Video Service Delivered over HTTP Adaptive Streaming

For the popularity of delivering video service over HTTP Adaptive Streaming (HAS) in the mobile network, developing an automatic method to evaluate customers' quality of experience (QoE) for HAS video service in real time is highly desired for network operators and content providers. This paper proposes a novel QoE evaluation strategy for HAS video service based on the specific features of HAS and data-mining technology. Via capturing the media description files of HAS and the request information of clients, the QoE can be monitored in real time for operators. The evaluation algorithm for the proposed strategy is trained and tested based on the data sets collected from subjective test results. The test results suggest that the predicted Mean Opinion Score (pMOS) measured by our evaluation algorithm has high correlation with the Mean Opinion Score (MOS) measured by the subjective test, and meanwhile the evaluation strategy is easy to implement for operators.