Study On Network Adaptable Techniques For Video Conferencing Systems

The video conferencing system is developed first by Bell Labs. Because it transmits audio and video mainly by analog signal method, it is very hard to process signal further. And also, none of its performance, price and conference cost has good popularity conditions. After the study of network techniques, multimedia techniques and encoding/decoding techniques had achieved tremendous progress and a series of international standard on video conferencing had been constituted sequentially, utilization of video conferencing system gets significantly progress.This thesis focuses on how to implement a video conferencing system that can achieve good audio and video effect at video conferencing system's application level though it is based on the network platform which can not provide quality of service.At first, the thesis analyzes deeply the characteristics of a video conferencing system itself. It indicates architecture, transmitted multimedia data's features and data frame's structure of a video conferencing system. Meanwhile, it also introduces relative important technology of video conferencing systems, including network status estimating and network bandwidth assignment and adjustment. Based on the analysis and study of important relative technology on video conferencing systems, this thesis brings up a video conferencing system which can be self-adaptable to its current network status, named for SAVC (Self-Adaptable Video Conferencing) system. At the same time, the thesis describes the system's design and details of implementation, including system's architecture, procedure and each function module.The system is composed of two parts: error control subsystem and communication bandwidth's adjustment subsystem. The error control subsystem adopts encoding/decoding method by the RS code to detect error and correct error. Also, it uses SARQ (Selected Automatic Repeat reQuest) method to send and receive data frames. The communication bandwidth's adjustment subsystem estimates the current network load status first, including available bandwidth, delay and jitter which are important arguments. Then based on real-time network status it can select properly audiovisual signal's quality level (for .example, video resolution, frame frequency etc.), so that it can provide a steady video conferencing environment for users.