Abstract
With the rapid advances on network communications technologies, video streaming has become a very popular multimedia application on the network. In the traditional multimedia system, constant bit rate (CBR) compressed technique is usually used to encode video because the CBR videos are appropriate for transmission. But this compressed technique results in video quality becoming fluctuant. In recent years, the variable bit rate (VBR) compressed technique is becoming more and more importance. This is because that the VBR video can provide smooth quality video. But VBR compressed technique results in high rate variability of the video and makes the video not appropriate for transmission on the network directly. An ideal video streaming mechanism should not only let user view smooth quality videos but also try to let videos transmit on the network smoothly. Therefore, we adopt VBR encoding technique to provide smooth quality videos to users and use bandwidth smoothing mechanism to reduce the bandwidth variation on transmitting these videos. However, majority of these researches for the bandwidth smoothing mechanism are based on the quality-of-service (QoS) network and can not be implemented on best-effort network. In this thesis, we propose a video streaming mechanism on the best-effort network. Our streaming mechanism provides congestion control to prevent network from congestion. Moreover, our mechanism also provides quality adaptation mechanism to adapt the video quality according to the available bandwidth. While adapting the quality of a video, our scheme also attempts to smooth the perceptual quality for the user. Then, our scheme computes a smooth transmission schedule in real time. According to this smooth transmission schedule, we can provide a multimedia service with the stable quality and the high efficiency for user.