The Design And Implementation Of The Wireless Audio-video Communication System Based On Embedded Linux

The rapid development of embedded and wireless communication technologies, especially thearrival of 3G era and the widely use of WiFi technology, provide an effective way for embeddedwireless audio and video communication. However, for average users, 3G video calls sets up twothresholds to them: one is high-end phone, the other is expensive tariff. In this paper, a low-costaudio and video communication solution is proposed based on embedded Linux and WiFitechnologies. In this scheme, the hardware platform uses S3C2440A ARM9 development boardexternal RT2571W WiFi module; the operating system is embedded Linux2.6.29; the networktransmission strategy is the combination of wireless AP access and Ethernet transmission; the C/Sand P2P network architecture is adopted in this communication system, in detail, the interaction ofvarious of signaling (such as registration and login signaling, etc) between the Client and the Serveris completed in the C/S framework and the point to point transmission of the audio-video databetween the Clients is completed in P2P framework.The major work of this paper is to design and implement the architecture and software of theaudio-video communication system, including the server software and the client software. Theserver software, which combines application server together with MySql database server, providesthe authentication and storage of users’information and other functions. The client software, whichis the focus of this paper, is developed based on embedded Linux and Qt/Embedded utilizingtechnologies such as audio-video capture, compression encoding, synchronization and networktransmission, and finally achieves real-time synchronous communication of audio and video.The main work is summarized as follows:First of all, the video capture based on Video4Linux2 is implemented and the MJPEG videocompression algorithm used in this paper is studied. The MJPEG algorithm is improved accordingto the idea of image change detection and the theory of inter-frame compression. The improvedalgorithm can greatly increase the compression ratio while reduce the compression time under theassumption that the background doesn’t change or change slowly (for example, camera is fixedduring the communication).Secondly, the software compression scheme based on G.729B voice compression algorithmand the hardware compression scheme based on AMBE-2000 voice compression chip are studied.For the former, the voice data collection on the OSS audio programming interface is achieved andthe G.729B algorithm together with its standard C source code which is optimized to reducecomplexity is studied. For the latter, our work includes designing the AMBE-2000 voice module together with its driver and the implementation of its corresponding applications. After comparingand analyzing the two programs, the software compression program is adopted in the clientsoftware development.Thirdly, audio-video synchronization technology is studied in four segments: collecting,sending, receiving and playing of voice and video. The network transmission technology is studiedtoo, including implementation of the P2P data transfer between clients at different private netsegments based on UDP Hole Punching and formulation of a simple and easy communicationprotocol.Fourthly, we achieved an elegant visualization client software based on the above relevanttechnologies and Qt/Embedded graphics development framework.Finally, after transplanting the client software to the ARM platform and setting up a testenvironment, we achieve synchronization wireless communication of audio and video.