Based on End-user Experience of Video Network Model Estimation
XU Fang-qiang;School of Information and Control Engineering, Qingdao University of Technology;
[purpose]According to the experimental data, the relationship between the initial buffer delay and the carton duration ofthe user experience evaluation variable and the network side variable(initial buffer peak rate, average playback rate in the play-back phase, E2E RTT) are established.[method]Firstly, the multi-linear function is used to model with Matlab software, and the re-lationship between the initial buffer delay and the network side variable is obtained. However, there is no obvious linear and nonlin-ear relationship between the Carton duration and the network side variables, so BP neural network is used to train and express therelationship between them.[result]The functional relationship between the initial buffer delay and the carton duration ratio of the us-er experience evaluation variable and the network side variable is obtained.
【CateGory Index】： TP183;TP391.41