Study Of Adaptive LDPC-UEP Scheme For SVC Video Transmission

In recent years, as the development and progress of the computers, networking and multimedia, video communication service has been popular among people because of its vivid and picturesque characteristic. Also the advantages of video transmission in the network transmission are more and more obvious, and occupy an important position. Transmitting the signal that contains video information requires high network bandwidth. So in the very limited wireless network conditions, to meet the multiple applications and multi-level needs of terminal is an important direction of the video transmission.Scalable video coding (SVC) based on H.264/AVC extension is the latest international video coding standard. It provides scalable coding features of space, time and quality, and acquires higher coding efficiency than previous video coding standards. Combination of the characteristics of the transmission system and the scalable characteristics of the video stream plays an important role on improvement of the video transmission system performance and the user's receiving quality.This thesis first briefly describes the development of scalable video coding standard and the main techniques of temporal,spatial and quality scalability in H.264/SVC. Secondly, it introduces some typical erasure codes and provides the unequal error protection algorithm of SVC based on LDPC codes. Thirdly it gives different protection schemes of SVC codes, and does the simulation experiments and analysis. On the basis of the above analysis and research, this thesis develops the adaptive protection scheme based on LDPC codes for SVC video transmission: when the channel conditions are poor, we can reduce the number of sending layers of scalable video and increase the protection of the channel coding ; when the channel conditions are better, we can increase the number of sending layers of scalable video and reduce the protection of the channel coding , so as to make the users receive better video quality. At last the summary of this thesis is made, and the further improving work in the future is gived.