Study On Pragmatic Encoding And Decoding Of Turbo Codes

This thesis puts the emphasis on the practical coding and decoding of Turbo code, including its soft-in/soft-out decoding algorithms, and the choice of its interleaver. First we introduce the iterative decoding algorithms of Turbo code, the soft-in/soft-out decoding algorithms of Turbo codes derived from block codes and convolutional codes. Then we introduce a simplified decoding algorithm for Turbo codes based on convolutional codes. This algorithm greatly decreases the memory needed for decoding while keeping almost the same performance. After that, the thesis discusses the design of interleaver for Turbo codes derived from block codes and convolutional codes separately. Using Hamming code as example, the thesis examine the design of good interleaver for Turbo codes based on block codes, and address the bound of minimum distance of Turbo codes using Hamming codes as constituent codes. For Turbo codes derived from convolutional codes, using (37,21)RSC as example, the thesis develops the design of good interlcaver, and examines the influence of different design on the performance of Turbo codes. Using the ways developed in the thesis, the performance of Turbo codes can be improved greatly on the same conditions.