QoE Assessment Of Video Service In Living Lab

With the rapid development of information technology, video services have beenwidely used in many areas. As one important part of video services, video qualityassessment has been studied intensively. The existing methods of video qualityassessment mainly use the video compression parameters and network transmissioncharacteristics to assess video quality. As most communication services tend to beuser-centric today, QoE(Quality of Experience) becomes a major research theme.Because QoE is the overall acceptability of service, the research of the overall perceivedquality of user in a video service needs to take account of not only technologies, butalso the aspects of business, environment and user mentality.This thesis analyzes the factors influencing QoE in the business domain, contextdomain and social psychology domain firstly, and proposes a basic model framework ofvideo service QoE assessment. In the traditional video quality assessment, the procedureand environment of subjective experiment are strictly limited. In order to reflect the realeffect of factors influencing QoE, this thesis builds a real-environment-simulativeplatform to study QoE of video services in the living lab. The platform simulates theservice in reality by realizing a network VOD system, and automatically collects thedata about the technology, context and user behavior as the quantified QoE factorsthrough different approaches and techniques of artificial intelligence. According to thestatistical data from the experiments, the parameters of video and network, as well asthe behavior of the user are selected to be the assessment factors by analyzing thespecific effect of every domain on QoE. Finally, based on the traditional video qualityassessment methods, an integrated quality assessment model for video services isproposed and the specific effect of the user behavior is analyzed deeply.