Summary: | 碩士 === 國立中正大學 === 通訊工程研究所 === 100 === Recently, the developments of video streaming technology are interesting, especially in the network transmission. With channel noise and various requirements from the user, the issues of packet loss and different quality of broadcasting service are still open. The efficient technique to realize the video streaming must consider the recovery of lost data and scalability to satisfy the various demands.
In order to meet the requirement of error correction and scalability, Scalable Video Coding and Unequal Error Protection are considered in this thesis. To analyze the layer priority of quality scalability and temporal scalability, the statistics of the loss in different Network Abstraction Level Unit are calculated in the transport layer. Therefore the optimization of the best video broadcasting quality can be achieved in the wireless network. Both of the channel coding rate and the number of scalability layer would be decided by the proposed scheme. Finally, the experimental results show the proposed Optimal Extraction scheme with coding rate selection has improved the average transmission quality about 1.52 dB than QL (Quality-Layer-based Extraction). Considering the packet loss condition with the proposed Optimal Extraction scheme, unequal error protection has PSNR gain about 0.564 dB, compared with the equal error protection.
|