User Throughput-Based Quality of Experience Evaluation for In Flight Video Streaming

Satisfaction of end users is one of the most important factors for every service provider. Given that a lot of factors along with the network conditions affect the user's experience, the network quality given by the Quality of Service (QoS) parameters alone cannot provide an accurate measure of the users Quality of Experience (QoE). Today, most of the airlines have started providing in-flight wi-fi services, which allow the passengers to use various Internet services while on-board the flight, such as send and receive e-mails, stream video from diverse online service providers, and so forth. Statistics show that more than 50% of the passengers use the provided wi-fi service to stream video, therefore, their perception of the video service will be determining for the service provider's performance. One easy way to evaluate the perceived video streaming (i.e. QoE) is by estimating the frequency of stalls. In our paper, we propose an evaluation method based on the user throughput values to determine the number of stalls the user may have experienced during an adaptive video streaming service. Our algorithm is not dependent on any user device property and requires only one input, namely, the throughput values delivered to the user device, which can be collected at either the router or if available, at the traffic shaper.