摘要
Turbo码的译码算法大致可分为串行译码算法和并行译码算法两大类。串行译码算法如MAP、LOG MAP等的研究已比较深入。但并行译码算法 ,尚有许多问题有待探讨。研究了Turbo码的并行译码算法 ,将Turbo码译码和图论结合起来 ,利用Bayesian网络图模型描述了Turbo码的译码过程 ,基于模型使用Pearl的信息传播算法 ,建立了Turbo码的并行译码算法。并对所讨论的并行译码算法进行了模拟 ,模拟结果表明
The turbo decoding algorithms can be divided into two categories of serial turbo decoding and parallel turbo decoding.The serial turbo decoding algorithms,such as MAP,LOG MAP etc,have been thoroughly studied.For the parallel turbo decoding algorithm,however,there are still a lot to be studied.With the combination of turbo decoding with graph,the authors propose to describe the process of turbo decoding using Bayesian network model,Based on this model,a parellel turbo decoding algorithm is established using Pearls belief propagation algorithm.the simulation results have shown find that proposed parallel turbo decoding algorithm is superior to serial ones in terms of decoding performance.
出处
《南京邮电学院学报（自然科学版）》
2002年第1期26-29,35,共5页
Journal of Nanjing University of Posts and Telecommunications
关键词
TURBO码
并行译码算法
信息传播
Turbo codes
Parallel decoding algorithm
Information propagation
