QoE Assessment of HTTP Adaptive Video Streaming

Quality of Experience (QoE) is a crucial characteristic of any multimedia service and must be accounted for during the service development and planning stages. Nonetheless, given its subjective nature, it is extremely difficult to use analytical methods to estimate the average Mean Opinion Score (MOS). Traditional progressive multimedia streaming is a well researched topic with respect to QoE, however, modern streaming services relying on advanced adaptive video streaming technologies, with specific characteristics, have yet to have an all-encompassing method for QoE estimation, as research work tend to focus on only one, or a small subset, of the technology's aspects, such as the impact of buffering events, bit-rate change frequency, or initial playout delay. This paper proposes a model for determining the QoE estimate of a playback session of HTTP adaptive video streaming, encompassing its complete range of characteristics. Several key-metrics are extracted throughout the playback session, and then analyzed by an analytical method able to predict the consumers' QoE. A subjective QoE survey is conducted according to industry's best practices and recommendations in order to validate the proposed models. The obtained results show that both subjective and objective estimations produce similar results, hence validating the proposed model.