Research On QoE-QoS Correlation Model Of Network Video Service Based On Statistical Analysis

With the development and popularization of communication and network technologies,the traffic of network streaming video services has dramatically increased.In order to control and manage the network video stream well and to guarantee the Quality of Service(QoS)and Quality of Experience(QoE)of network video,it is very to model QoS parameters effectively.Based on real network video stream data and the MOS(Mean Opinion Score)calculation model based on the ITU-T P.1201 standard,this thesis analyzes in depth the PDF(Probability Density Function)of QoS parameters of transmission rate and jitter with regard to MOS values,and the statistical correlations among the three.Combined with the distributions of MOS values of video service,the study established a statistical model of MOS value on jitter and video bit rate.Compared with existing methods,this model can provide service providers and network operators with more accurate reference for setting QoS parameters under given user's QoE constraint,and thus has good application prospect.The main contents of this thesis are as follows:Firstly,it introduces the related concepts of QoE and presents some existing models of two major categories of QoE video,subjective quality assessment and objective quality assessment,and exlains the principle,advantages and disadvantages of these models.Then,this thesis introduces the evaluation model of video QoE under the ITU-T P.1201 / Amt2 standard in detail.According to this model,through the simulation of video playback algorithm,we have statistics in the initial buffer time,the number of interrupts,interrupt duration and video bit rate,and calculate the MOS value.Then the paper studies the jitter and the probability density function of the video bit rate in each MOS value range of standard definition,high definition and ultra definition online video streams.Finally,for the three diff resolution video streams,this paper establishes the relationship between QoS parameters and MOS values.In the case of statistical analysis,it is shown that for given MOS values can be more accurately obtained QoS parameters(delay jitter,video rate)value range,and as a reasonable basis for the deployment of network resources and for the quality of user experience optimization.