Research On Adaptive Control Scheme Of Streatming Media Buffer Playout

As the development of the network communication technology, the current network not only can offer a simplex voice service as before, but also can offer multi-media service which integrated by voice service, data service, and streaming media service. Among the multi-media service, the streaming media service is the most important service, and it is used more and more widely in the network.With the wide usage of the streaming media service, the users also have higher requirements on the quality of the streaming media service. In the transmission process of streaming media service, the requirements of the timeliness and the data volume should be ensured. However, in recent years, plenty of researches demonstrate that the service flow which is transmitted on the network has self similarity character in a large time scale. The character will lead to unexpected burst character of the service flow. That is to say, at some time section, if the average flow value is big, in the next time section, the average flow is likely to big. At some time section, if the average flow value is small, the next time section's average flow value is likely to small. The requirements of timeliness and the data volume cannot be ensured with existence of the self-similarly character. The play quality of the streaming media service will be affected seriously. The existing self-similarly character is the important factor which has influence on the play quality of the streaming media service.Therefore, in order to ensure the transmission quality of the service and the play quality in the terminators, many adaptive buffer play schemes are produced. But, the current control strategies and technologies do not fully consider the self similarity character which can breed the data burst character in the network. The character will make the current existing schemes cannot play a good effect, cannot smooth the delay jitter and restrain buffer overflow, even cannot ensure the play quality of the streaming media service and a nice user experience in the terminal.Considering the existence of corresponding schemes cannot adjust the self similarity character, this thesis proposes the below two main tasks.Firstly, design an adaptive play speed adjustment scheme based on dynamic double thresholds. On the basis of the original schemes, this scheme can better adapt the changes of the network's service flow character. Using the introduced dynamic double thresholds, this scheme can better smooth the delay jitter which results from the network burst because of the existing self similarity, decrease the variance range of the play speed, maintain the occupation rate of the buffer a stable status, realize the adaptive play of the streaming media service, ensure the timeliness character of the streaming media service and enhance the play quality of the streaming media service.Secondly, design a dynamic buffer adaptive control strategy based on reservation resource. The existing self similarity character will increase the probability of the overflow and underflow of the buffer. The buffer allocation mode of the strategy is designed as a dynamic mode instead of the original static allocation conception. The reserved buffer can save the overflow data. The scheme can ensure the data volume of the streaming media service quality and enhance the play quality of the streaming media service. Meanwhile, the dynamic buffer can enhance the resource utilization rate, adaptive the change of the service flow and decrease the frame loss rate.Finally, this paper implements experiments for the two designed schemes. The results show that the adaptive play speed adjustment scheme based on dynamic double thresholds can better adapt the changes of the network status, smooth the delay jitter of the network, and make sure the timeliness character of streaming media service. The results also show that the dynamic buffer adaptive control strategy based on the reservation resource can reduce frame loss rate when the network's burst results in the overflow of the buffer, and also can ensure the data volume of the streaming media service.