A Model for Mapping Combined Effects of Quality of Service Parameters and Device Features on Video Streaming Quality of Experience

Abstract

Maintaining quality of streaming video is challenged by network faults resulting into freezes and rebufferings on the video. On top of the network effects, device features have impacts on the image of the video frames displayed during streaming. Despite the simultaneous impacts of video quality from network and device, previous studies considered individual impact of network parameters or devices as influencing factors to propose Quality of Experience (QoE) models. This study proposed QoE model by mapping combined effects from both network and device on video streamed QoE. An experiment to study the effects of video quality from combined effects of network and device over the wireless involved 35 subjects. Combination of packet loss, packet reordering, and delay were emulated using network emulator following Design of Experiment methodology. Through analysis of variance, the study found that packet loss had the highest impact, followed by device features, reordering, and delay on the video QoE. From the combined effects, two-way interactions and three-way interactions had significant effects on the video QoE. Through additive and linearity behavior of the input factors from network and device on video streaming QoE, a multi-factor model was derived.

Keywords: Design of Experiment (DOE); Mean Opinion Score (MOS); Quality of Experience (QoE); Quality of Service (QoS); Video Quality Assessment