Research On Quality Of Experience Collaborative Evaluation Mechanism For Relay-based Multipath Transmission

Internet "Best-effort" transmission mechanism is difficult to guarantee the Quality of Experience(QoE)of "real-time conversational class service" such as network audio and video phone,which has seriously hampered the network service innovation.Relay-based multipath transmission is an optimized transport solution to application layer network resource scheduling,constructing a plurality of incomplete intersection transmission paths between media terminals through application relay service.This transport mode not only can expand the "bandwidth" between the terminals but also can further reduce the influence of the network mechanism defects such as "random congestion" and others through the multipath scheduling and redundant transmission.And then,qualified quality of experience of the media transmission will be achieved."The service quality of experience" is the required criteria of network resource scheduling and multipath transmission control."QoE real-time evaluation" is the decision basis of "relay path generation and selection" and "end-to-end multipath transmission control(payload distribution)".The traditional single path transmission mode only observed the effect of network transmission of QoE.And QoE is determined by the path quality of service(QoS)such as related to the bandwidth and throughput,network delay,packet loss rate,jitter and other network conditions.But in multipath transmission mode,any independent paths in QoS conditions cannot determine the final service QoE.And multipath transmission QoE is the product of "media service characteristics","each path QoS condition of the path set","multipath transmission control method" and other factors.The transmission related service QoE researches mainly concentrate on the media service characteristics,single path QoS condition model and analysis at present.While service QoE quantitative models and evaluation methods of multipath transmission are still blank.The overlay network multipath transmission technologies,media transmission service models and end-to-end QoE evaluation systems are revicewed.Combined with real-time conversation media service multipath transmission scene feature,the decision demand of "relay path selection" and "payload distribution',this article researches from the parts of service quality of experience quantitative evaluation model,collaborative evaluation methods and policy decision model of relay-based multipath transmission.The main contents and innovations include:(1)Scene Constraint real-time video service QoE quantitative evaluation model and multipath coupling reasoning evaluation method are proposed,which mainly inspect influence of "service features" and "different path QoS conditions" for service QoE.This method takes the coupling relationship between network transmission QoS parameters of the multiple paths into account.According to defined service rules and process of QoE evaluation,this method gives user experience grading calculation rules and corresponding transmission control guidance rules of media service.At the same time,it uses forward and reverse hybrid reasoning model to evaluate service QoE model,and ultimately carry out service QoE evaluation.(2)QoE collaborative evaluation method based on fuzzy clustering heuristic algorithm is proposed which uses big data processing method of analysis historical evaluation data in accord with network status performance perception and parameters developed service QoE evaluation rules.It takes advantage of heuristic evaluation rules and fuzzy clustering analysis method to build fuzzy clustering function for evaluation and analysis of sampled data.Based on the similarity matching and clustering evaluation of historical evaluation data,a fuzzy clustering heuristic algorithm based on big data analysis was formed.And the evaluation method of service QoE is evaluated by it.This method avoids large errors for uneven coverage of the existing evaluation linear law,gives accurate service score information and reduces the degree of deviation from the evaluation according to parameters difference features of the evaluation sample.(3)A kind of QoE collaborative evaluation mechanism for relay-based multipath transmission is proposed.The mechanism contains two network components,named QoE monitoring and evaluation management client,QoE monitoring and evaluation management server.Wherein,QoE monitoring and evaluation management server contains four modules,including the parameter filtering module,media database module,service QoE scored calculation module and monitoring and management module.QoE collaborative evaluation mechanism uses "centralized" evaluation management mode to evaluate media service.Centered on QoE monitoring and evaluation management server,QoE monitoring and evaluation management client real-time feedback evaluation parameter information of media service requires such as node location,network transmission QoS parameters,user subjective parameters,assists the server to complete the QoE evaluation management.QoE collaborative evaluation mechanism helps to generate and apply historical data from a global perspective,and avoids the problem that could not be evaluated which due to lack of database.This method reduces the calculation pressure of the client,and enhances the accuracy and efficiency of QoE evaluation feedback in media transmission.(4)VP9 real-time high-definition video session service QoE evaluation policy decision method is proposed.In view of the multiple transmission schemes and corresponding transmission requirements,this method designs service QoE evaluation policy decision rules and implementation scheme.To ensure the quality of service of the transmission media,QoE evaluation policy decision method designs the rules of multipath hybrid redundancy transmission to adjust the media transmission control,and ultimately to meet multipath transmission control and user experience requirements.QoE evaluation policy decision method ensures the quality of service of the media,and enhances the reliability of service transmission.Two goals of the service QoE collaborative mechanism are to establish service QoE evaluation method and give service QoE evaluation policy decision guidelines.QoE evaluation method provides a theoretical basis of the end users and the server.While QoE evaluation policy decision is to select the optimal transmission scheme which is suitable for media transmission control,and regulates the media control according to policy decision rules.The research results show that the service QoE collaborative evaluation mechanism can provide accurate user quality of experience score for the transmission media service,and optimizes the transmission control of media service.