Design And Implementation Of Speech Coder And Multimedia Synchronization Based On H.323 Protocol Stack

The rapid development of techniques on multimedia networks and communications has greatly improved the environment and driven the advancement of multimedia applications, which result in the mass emergence of multimedia communication systems, such as multimedia conferencing system, Video on Demand and so on. These systems have provided many kinds of point-to-point multimedia communication services on IP networks. Speech processing and synchronization in point-to-point video conferencing will be discussed in this dissertation.There are many kinds of speech coding algorithm, the most extensively used among which is Code-Excited Linear Prediction (CELP) coder. They introduce the techniques of vector quantization (VQ) into speech coders, and generate very good speech quality in a wide bandwidth of medium and low bit rate. However, their complexities are very high,and searching their VQ codebooks is always time consuming. G.728 will be introduced as a example of application of CELP. In order to reduce its complexities and improve its executing efficiencies, we will optimize the algorithm based on processor and based on the theory of G.728.The video conferencing needs to support real-time speech and video communications,but they are transmitted on IP networks with no (Quality of Service ) QoS guarantee, the relation between audio and video on time is corrupt because the packet lose or jitter is taken place in the during of transmission, which will degrade their playback quality. Media stream synchronizations are analyzed and studied, the implementation schemes of the related synchronization control are given in this dissertation.