| Along with the rapid development of mobile communications as well as deep-space communications, the applications of multimedia data transmission over the wireless channel become significant. The design target of the next generation wireless communication system is to implement wideband and high-speed data transmission in high-speed mobile environment. Because of the limited bandwidth resource, the high error rates and the severe fading over the wireless channel, there are many challenges in order to ensure the reliable multimedia data transmission. Though international image/video coding standards have provided solutions for effective data compression, the compressed bit stream is very sensitive to even a single binary error due to the characteristic of the VLC (Variable Length Coding). Therefore, it's urgent to study the high reliable channel error protection technology in order to improve the transmission quality from end to end.As a new kind of channel encoding method, turbo coding has been attracting many academic attentions for its perfect decoding performance near the Shannon limit. It has been applied by the 3G mobile communication standards and the deep-space communication standards. On the other hand, the practical spacial communication systems are usually heavily restricted by delay and complexity, so that the classical Shannon separate coding systems cannot achieve the optimal performance, nevertheless, joint source-channel en/decoding technologies obtain increasing research interests in recent years. For above considerations, joint source-channel en/decoding based on turbo codes are selected as the research theme. Based on the Shannon information theory, traditional source compression coding and channel error correction coding methods are introduced. The essential technologies related to turbo codes are studied, including the turbo principle and its applications, bandwidth efficient turbo coded modulation, and joint source-channel en/decoding methods combined with turbo codes. Especially, we present the LCB iterative stopping algorithm, the symbol-based SOVA algorithm suitable to the TTCM system, and two joint source-channel decoding algorithms. An application of the new method to the image data transmission is also discussed.The main work and contributions are:1. To the large delay of turbo decoding, a rapid turbo decoding method and the LCB (LLR Characteristic Based) stopping algorithm are proposed. Without losing the BER (Bit Error Ratio) performance, the new algorithm can effectively reduce the average iteration number, providing a solution for applying turbo codes to the real-time business.2. To the issues of the limited bandwidth resource and the severe fading over the wireless channel, bandwidth efficient coded modulation methods for turbo codes combined with the TCM (Trellis Coded Modulation) technology are studied. A symbol-based SOVA (Soft Output Viterbi Algorithm) suitable for the Turbo TCM system is proposed, which can reduce the decoding complexity largely.3. Based on residual source redundancy of the variable length coding, a joint source-channel turbo decoding method with source a priori information is presented. By constructing a bit-level super trellis with compound states, bit transition probabilities of the VLC tree can be utilized in the iterative decoding algorithm, which decreases the symbol error ratio of variable-length encoded data.4. Combining the VLC with turbo codes, a new hybrid concatenated joint source-channel coding model is proposed, which is named as variable length turbo codes. A symbol-level joint trellis is obtained by means of sub-state merging, and a variable-length symbol APP algorithm (VLS-APP) is implemented, which is more suitable for the source a priori characteristic. At the decoder side, the system transmission performance can be improved by joint source-channel iterative decoding.5. Based on the CCSDS IDC (image data compression) standard, the high-fidelity scalable compression algorithm suitable for the spatial communications is studied. In addition, an application of the proposed variable length turbo coding to the embedded image bitstream transmission is also discussed, which benefits to the image reconstruction quality.
|
PDF Full Text Request