Algorithms and software tools for distribution of multimedia data streams in client-server videoconferencing applications
Anton Igorevich Saveliev, Andrey Leonidovich Ronzhin · Pattern Recognition and Image Analysis · 2015
Most videoconferencing applications are based on the client-server architecture of processing and transmitting multimedia data among the participants of a conversation. In this paper, different methods of distributing and processing data streams are analyzed and optimized with the aim of reducing the load on the server-end portion of the videoconferencing system, as well as on client devices. First of all, the analysis of the following phases of data processing: generation and removal of audio and video streams, server/client data exchange, organization of interoperation of stream chains, and storage of and search for data streams in the server-end portion. After optimization, new algorithms were developed for data stream processing for a videoconferencing application rivaling the segment’s top software, which is Skype. The use of random access memory (RAM) and CPU time was reduced tenand twofold, respectively.