Research Of Audio And Video Session Technology Based On SIP Protocol

With economic development and the improvement of living standards, accelerated the pace to transfer part of health resources to the community, gradually improve the community health service institutions and resource allocation of hospitals, but also gradually increase the supply of community health service ability. The purpose of this study is to design an audio and video session system, based on community health service, used in local area network of within community or between the communities.Before the narrating of research project, this paper first introduces the background and research significance of the topic, and the research status of the SIP and audio-video synchronization, and then the general theoretical knowledge of SIP protocol is summarized. Audio-video synchronization and SIP combined with P2P technology are the two mainly focus of the article research. The Audio-video synchronization based on time stamp and audio-video synchronization coding based on variable code length are th two research points of audio-video synchronization. The implementation of audio-video session system is divided into five modules:capturing, coding, network transmission, decoding and playing.(1) Audio-video synchronization scheme based on timestamp is implemented around the whole system framework, in other words, every module of the system will use the scheme to control synchronization. In the capturing module, audio and video data are capturing respectively, at the same time, the capturing starting time of every frame of video and every block of audio data would be recorded, and the time is called time stamp, the relevant data is divided into synchronous data block according to the rules of synchronous data block. In the following module, the input and output data is always a synchronous data block, it ensures that a synchronous data block, reaching playback module, is synchronously played.(2) The technology scheme of audio-video synchronization coding based on variable code length inputs audio encoded data, each read2or3bits of data as a set of variable length code group according to the rules. In the stage of video codec standards H.264inter prediction, the mappings between various variable size blocks and the data of code groups were based on formula. The encoding side, mapping the encoded audio data variable length code group to variable size macro block partition mode, and the decoding side, extracting the corresponding encoded audio data according to the mapping relationship.(3) Mainly considering both shortcomings of SIP and P2P, in order to make up for their shortcomings, adopted the combination mode of SIP and P2P, and then to become a completd architecture. Although the SIP protocol itself has the characteristic of P2P, but this characteristic is dependent on SIP server on the Internet environment, the implemention of P2P technology need help of server, that is in keeping with the system application background. The backend center of community network can set up the server program (The user agent server, defined in the SIP protocol, is a entity element), in this way, it can not rely on SIP server, and the server program can control the entire community network terminals, that can shoot two hawks with one arrow.In terms of the combination of SIP and P2P technology, a major difficulty is to solve the NAT traversal, The NAT type is mainly divided into symmetric and asymmetric NAT, due to the lack of time and the more complicated of symmetric NAT traversal, only achieve the asymmetric NAT traversal now, so the next step in the research is focused on symmetric NAT traversal.