Master Thesis - 2009

In recent years, video streaming services have become popular and major applications in high speed networks. A video content is composed of some pieces of data which have various importances. The video can be played back with acceptable quality even if data of relatively low importance are lost. On the other hand, quality degradation due to loss of important data propagates during certain term. In order to recover the lost data, a lot of methods have been studied. As one of these methods, a packet recovery method making parity packets from data of multiple streams has been proposed. Even if one stream is suffered from many lost packets, this method can recover the lost data by decoding parity packets with data of other correctly received streams. However, this method does not consider the various importances of video data. Considering the importance of data, video quality can be further improved.

This study proposes a recovery method using multiple streams with considering importance of video data. The proposed method gives priority for recovery to each data of multiple streams based on both inherent importance in intra-stream and loss probabilities in inter-stream. This method is designed for recovering important data with higher probability. In addition to this, this method deals with the data of less received stream as the important data. Moreover, the proposed method decides appropriate number of parity packets. This is because the transmission of too many parity packets might affects the other traffic, and that of too fewer packets might recover insufficient amount of data. The proposed method can improve total video quality of multiple streams.

The proposed method is evaluated through computer simulation. The simulation results show that the proposed method can recover more important data with higher probability. Furthermore, the proposed method can also achieve almost the same restoration ratio among streams of recovery target.