Research On The Simulations Of Sub-block Parellel Turbo Decoding Algorithms And Design Of Collision-Free S-Interleavers

Turbo codes, which employs iterative decoding algorithm of soft input and soft output has excellent error correcting performance. However, with the high complexity of the decoding algorithm and iterative decoding structure, the decoding delay of turbo code is large that restrict its application for high-rate and real-time communication systems.To solve the problem of decoding delay and improve the decoding throughput, many scheme have been proposed, one of the most effective way is to use sub-block parallel decoding structure that each frame data is divided into several sub-blocks with the same length, and each sub-block is decoded by child decoder simultaneously. The more of the sub-block is divided or the shorter of the sub-block is, the smaller of decoding delay and higher throughput rate is. However, additional problems are introduced into turbo decoding by employing block parallel decoding structure:boundary recursive variables of sub-block need to be accurately initialized, if not, the error correcting performance of turbo code will be loss; due to the randomness, interleavering and de-interleavering of the parallel block decoding data will be more complex, and memory contention may emerge to increase the decoding delay, so collision-free principles need to be considered in the design of parallel interleavers. This article is mainly to study and analysis the performance of sub-block parallel decoding algorithm for turbo code and design the algorithm for generation of collision-free S-random interleavers.To get a preliminary understanding for theory of turbo code, a brief introduction of turbo encoding and decoding fundamentals is given at the beginning of the thesis, then the error correction performance of Turbo code of the3rd generation partnership project long term evolution (3GPP LTE) standard, with different condition and different kind of interleavers, is investigated by simulation of turbo decoding. Subsequently, parallel decoding structure of turbo code and several kinds of existing decoding algorithm is present. Then the simulation of turbo decoding for different algorithms is given, the performance of a combined algorithm is compared with them. The applicability of the combined algorithm is pointed out. Finally several existing design scheme of collision-free (CF) interleavers is present, and a new algorithm for generation of collision-free S-random interleavers is proposed, which can efficiently transform any semi-randomly CF interleaver to be S-random for de-interleaving operations, where the spreading factor S can attain its upper bound while the interleaver keeps collision free property for interleaving operations. Due to the large spreading factors, this kind of CFS interleavers can result in excellent turbo decoding performance with original S-random interleavers which is proved by turbo decoding simulation.