Research On QoS Optimization Of Video Conference Transmission And Video Conference System Design And Implementation

With the development of communication technology,the information exchange tools between people have evolved from original texts and face-to-face conversations to more accurate and rapid ways,such as making phone calls and sending text messages.Among them,video conferencing is one of the more widely used means of remote information exchange.To enhance the user experience of video conferences,it is increasingly important to study how to improve the quality of service(Qo S)of video conferences.This thesis mainly focuses on the research of video transmission and video transmission congestion control mechanism in the video conference system.Firstly,this thesis analyzes the transmission scheme and related transmission protocols of the video conference,and studies the BP neural network,RNN neural network,and GRU neural network.According to the temporal characteristics of the video traffic of the video conference,this thesis integrates the idea of wavelet transform into the cyclic neural network,and proposes a video traffic prediction model based on the GRU neural network after the data is processed by the wavelet transform.The data is processed by the wavelet transform and decomposed It is the high-frequency and low-frequency components,and the GRU neural network model is used to predict the decomposed high-frequency and lowfrequency components,and the prediction result is reconstructed by wavelet transform.The result after reconstruction is the video traffic sending situation in the future.The experimental results show that the video traffic prediction model proposed in this thesis can more accurately predict the data at the next moment.Secondly,this thesis puts forward an improved TFRC congestion control algorithm;the estimated value calculated by the original TFRC algorithm and the predicted transmission rate reference factor are used together to adjust the transmission rate to adjust the video traffic transmission rate.Experiments prove that the improved congestion control strategy proposed in this chapter can better guarantee the Qo S of video transmission.Thirdly,this thesis designs and implements an SFU-based mobile multi-person audio and video conference system based on the requirements of the actual project,and conducts performance testing and analysis on the system's memory,traffic,FPS and other indicators.The test result shows that this system has low memory usage,normal traffic conditions,and stable video FPS conditions.The actual application of video conference shows that the system has robustness and stability.