|
With the rapid development of digital communications, computer networks and the universal application of Internet technology, multimedia communication has become an inevitable trend. The efficiency of video information with certainty, intuitive, superior performance, video transmission in the network transmission advantages more and more obvious, occupy important position, however transmission contains video information signal needs higher network bandwidth. Therefore, in very limited wireless network conditions, maximize meet the needs of the terminal a variety of applications and multi-level, is an important direction for video transmission. Look from the perspective of information theory, Shannon's source coding and channel coding separation theory optimal encoding performance can not be done in practical use, in fact, the separation theory limits the system to achieve optimal performance. Later, people began to study the source / channel joint encoding (Joint Source Channel Coding, JSCC) can considering two aspects of the source and channel conditions, in order to achieve the overall optimization of the transmission performance. JSCC by optimizing between the different parts of the source of the allocation information, the bit rate between the source and channel, so that network bandwidth is limited, an end-to-end transmission of minimum distortion. Unequal Error Protection (Unequal Error Protection, UEP) is a research direction of the source / channel coding. The scalable video stream layers having different levels of importance, the use of unequal error protection techniques to improve the error recovery capabilities are feasible. Specifically, the more protection bits assigned to the more important frames and layers, so as to have a greater robustness. So, encoded according to the video signal source, the data generated the different importance of an unequal error protection of source / channel coding method. The main content of this paper include: first introduced wireless video transmission technology research purpose, meaning and Research, a of H.264 layered video coding technology to discuss the fine-grained layered video coding algorithm. Then, starting from the information theory, Shannon separation theory, analysis of the superiority of the source / channel coding in a communication system, the rate-distortion theory rate allocation. Proposed scalable video coding (Scalable Video Coding, SVC) source / channel coding scheme, the simulation tests on the algorithm in JSVM7.10 procedures. In this paper, unequal error protection of the source of a low density parity check code (Low Density Parity Check Code, LDPC) SVC / channel joint coding scheme. In the proposed scheme, a group of pictures (Group of Picture, GOP) as the basic unit, first by calculating each layer rate-distortion function DR envelope curve of each frame, then pull through joint rate allocation algorithm using binary search constantly adjust The Lagrange multipliers optimize limited allocation of bits to each quality layer, so that the conditions that satisfy the bit rate constraint, the end-to-end distortion is minimized. The complexity of the proposed joint coding algorithm to produce fine-grained enhancement layer video streams at any bit truncated, and have strong error resilience and adaptability channel. The paper gives a source / channel the joint coding transmission system diagram, and briefly describes the basic functions of the various parts. Finally, the simulation experiments, the performance graph in the redevelopment. The results show that the method according to the different conditions of the channel bandwidth and packet loss rate, rational optimization of the source rate and channel bit rate, in order to achieve the minimum end-to-end distortion. By comparison, the proposed scheme has made a good reconstruction video quality, at the same time ensure better quality video while in low complexity.
|