Turbo Decoding and its Application Based on CUDA Parallel Computing

• Main Authors: Shin-hung Lin, 林信宏
• Other Authors: Shou-Sheu Lin
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/v9mu8a

碩士 === 國立高雄第一科技大學 === 電腦與通訊工程研究所 === 103 === Turbo code is a high performance error-correcting code, and it has been widely used in various communication systems currently, including long term evolution (LTE) standard for 4G mobile communication. Since the turbo decoding needs to use continuous iterative decoding to reduce the error rate, it leads to quite a large amount of computation and spend a lot of elapsed time. We use the highly parallel architecture which is graphics processing units (GPU). Turbo code could be parallelized and implemented on the Compute Unified Device Architecture (CUDA) platform. On the other hand, we simulated three kinds of stopping criterion. Respectively Genie, Cyclic Redundancy Check (CRC) and Hard Decision Aided (HDA). Finally, we use the simulation to compute number of iterative to analysis the pros and cons of stopping criterion.