Quality Of Experience Evaluation Of OTT Video Streaming Service

With the continuous advance of the tri-networks integration,OTT business has become a new developing trend and technology direction.And OTT video streaming service,owing to its huge demand,turns to be the focus of attention in the Internet.So it is increasingly important to improve user-perceived quality and ensure their satisfaction.Different from the traditional concept QoS(Quality of Service),QoE(Quality of Experience)is user-oriented,indicating the overall acceptability of the service.Therefore,the study of QoE assessment for video streaming service is a hotspot in the field of network management.Firstly,QoE assessment considering the effects of the pause position is put forward.This paper,starting from the progressive download technology,analyzes the characteristics of the mainly used HTTP video streaming and its key factors.The author focuses on the rebuffering events over the video streaming,and based on the three classic application layer performance metrics(initial buffering time,rebuffering frequency and mean duration of rebuffering),proposes two new metrics:the pause location and mean time interval of pauses.Through the mathematical derivation and simulation experiments,this paper studies how these pause position factors affect QoE,and uses the Back Propagation Neural Network(BPNN)to establish the mapping model from the application metrics to QoE.Secondly,in order to get more close to users,the author provides the QoE evaluation method of HTTP video streaming based on the user interactive behaviors,which is in the light of real-life application scenarios of the service.User interactive behaviors play a dual role during the video streaming:reflection and effect.Combining application layer performance parameters with user behaviors,the QoE metric system is built.Thus,this paper designs a comprehensive and in-depth evaluation model.The results from experiments show that QoE assessment is improved with the addition of user interactive behaviors,leading to better personalization.In conclusion,the proposed QoE evaluation methods of OTT video streaming in this paper,have high accuracy and good performance.Finally,the expected goal of this study is achieved.