Research And Implementation Of Adaptivemediasystembased On Converged Network

As network develops, people use more and more multimedia applications in their everyday life, like multimedia conference, multimedia call and remote medical treatment. But video transmission requires more bandwidth compare to voice and text transmission. So when people are making video calls over Internet, Network connection quality will have a greater impact on video quality.However, Current converged network does not provide any quality assurance functional entities which means there isn’t any guarantee of user experience when people are in a multimedia conversation. Therefore, this paper is aimed to research and implement a multimedia adaptive system to provide a good Qos and QoE in multimedia calls.This paper firstly introduced the technologies related to the system, including SIP protocol, RTP/RTCP protocol, and introduced Java-based implementation framework. Then the needs of media adaptive system is analyzed, including Its location in the network, working scenarios and functional/non-functional demands. Secondly, we researched the mechanism of adaptive media stream transmission, including how to choose the transmission protocol, how to calculate the variables which represented the network status, and the basic principles for media adapting. This paper proposed two new adaptive transmission strategies:l)if the packet loss rate is too high to maintain a normal conversation, and users could hardly communicate normally, system will automatically switch the transmission medium, which will change the video stream to audio stream. Because audio stream usually take up less network resources.2)if the network encountered an interruption, and the endpoint luckily have a back-up endpoint, the system would automatically switch the session to its backup endpoint. Finally, based on the above two strategies, this paper designed and implemented a media adaptive system. The system could monitor the status of current network, when the network status is becoming too bad to support a smooth conversation, the system would make the most appropriate adaptive strategy according to the current network. The system has been tested having a good outcome, so it has achieved the design goal.