Research On Multicast-based Multimedia Communication With Adaptability

Thanks to the development of multimedia technology, people can integrate different kinds of media such as video, audio, text, and image into their applications. This trend has improved the human computer interface greatly and changed the way people working with computer. In recent years, multimedia application system has been widely realized in video on demand (VoD), video conference, distance education, tele-medicial and digital library. We believe that multimedia application will affect people's work, living and entertainment deeply in the future. Multimedia communication in the network has some basic feature: high bandwidth usage and massive data transportation. In order to solve these problems, multicast is used to transmit digital media in the Internet to save bandwidth and utilize the network efficiently. But there exists two problems: â‘ The ability to adapt distributed network's heterogeneity is not good. The reasons that cause this problem includes: the rate and the performance of computer differences between users. The heterogeneity problem made the efficiency of multimedia communication is low. â‘¡Adaptability to the network's dynamic change is not good. In practical applications, the available bandwidth and packet loss rate change randomly. But today's multimedia communication system did not take effective measure to deal with this problem. To solve these problems, in this paper we first analyze the basic multimedia communication protocol: Real-time transfer protocol (RTP) and Real-time transfer control protocol (RTCP), then make research on the Receiver-driven Layered Multicast (RLM), finally provide an multicast-based multimedia communication model with adaptability. The design goal of this model: first, the video and audio data are layered at the source and transmitted to the multicast environment. By selectively joining or leaving different layers according to their available bandwidth, receivers can achieve the best quality signal that the network can deliver. Second, session manager processes all receiver's feedback information to monitor the network's status, then changes the sender's transmission rate dynamically. Third, realize synchronization among video and audio data using RTP and buffer mechanism at the receiver. The results show that: (1) the model can solve the heterogeneity problem. (2) the model can lighten the network congestion to some extent because its feedback mechanism. (3) the model can synchronize the presentation of multimedia when it transmit live video and audio stream simultaneously. At the end of this paper, we use Java Media Framework (JMF) and RTP technology to implement the model's some function, mainly the receiver and sender parts.