Construction Of Mobile Terminals Application Platform And Research Of Congestion Control Algorithm Based On Video Transmission

With the popularity of smart mobile terminals on a global scale,and rapid developments of wireless communication technology,data coding technology and streaming media technology,real-time video streaming application has now become the main trend in the field of video transmission,which attracts the attention of various fields,such as the military,business and politics.Real-time video transmission based on mobile terminals makes video conference,video monitoring and video broadcast possible at anytime and anywhere,which brings great convenience to people's work and social life.Congestion control is one of the most important algorithms in the field of video transmission.Good congestion control algorithm can effectively improve the performance of video transmission,meet high bandwidth demands and solve the problems of streaming media,such as packet loss and high delay.Considering increasing demands in fluency,clarity and diversity of video sent and received by mobile terminals,researches on video transmission application platform and congestion control algorithm become more and more significant.This paper studies these two aspects,and the main work is as follows:Firstly,an application platform based on video transmission with mobile terminals has been built.It completed functions like video broadcast,video-on-demand and related social applications.The platform was developed in Java,with application server named Tomcat,and streaming media server named Wowza Media Server.From the perspective of video transmission and social applications,the platform accomplished main functions,such as live streaming,video-on-demand,login and registration,friends following,fans removing and video comments.In this paper,the functional and non-functional requirements were analyzed and confirmed at first.After that,the design of the overall architecture and working flow of the whole system was described in detail.Then,from the video acquisition terminal,server side and video playback end,modules implementation of the video transmission application platform were made respectively.At last,the video acquisition subsystem and video playback subsystem were completed.Video acquisition subsystem mainly completed video collection,upload and video personalized settings,including name,setting the image as the theme,setting permissions,etc.Video playback subsystem focused on receiving,watching and comments of the video.Secondly,this paper proposes a congestion control algorithm based on Partially Observable Markov Decision Process(POMDP)in video transmission,which can minimize the distortion of streaming media and improve the quality of video transmission.The implementation of the algorithm includes the following aspects.First,we analyzed the model of unicast video transmission.With the POMDP model,the algorithm evaluated and updated the state situation of network congestion with history information of the system and the current situation observed by users,and calculated the optimal updating strategy of each user.Then,we researched and analyzed the way to solve the optimal updating strategy,and put forward Point-Based Heuristic Search Policy Iteration Algorithm(PBHSPI).The simulation results show that the proposed congestion control algorithm based on POMDP can improve the quality of video transmission in wireless network.