Layered Audio Codec Based On Sinusoidal Model And Reserach On Its Application

Currently, there are two popular ways to assure quality of service (QoS), one is supplied by network provider who classifies customers and gives them different services. Other is from media application system provider who uses acknowledge packages to adapt to the network situation. But as we know, in first method network resource cannot be used effectively, and the greatest difficulty for second one is that terminals cannot get accurate information about network traffic.So, this thesis purposes a new idea to improve the quality of service (QoS) based on one special kind of codec in terminals. In this method, middle nodes of network can adapt to situation of network traffic actively. According to this, this thesis analyzes several current technologies on speech and audio signal processing, and a simple layered audio codec based on sinusoidal model (LACS) is designed and implemented.Based on LACS, thesis analyzes the applications of layered codec, including audio and video codec, in current IP QoS assurance mechanism to improve QoS. A potential method to use layered codec is identified, simulated and studied in this part.At last, this thesis makes some investigation about application of LACS and layered codec in video&voice over IP (WoIP) system. For several components of H.323 visual multipoint meeting system, how to use layered codec is mentioned in this part meanwhile.