基于平均熵的Turbo码迭代译码停止准则  

Average-Entropy Based Stopping Criterion for Turbo Iterative Decoding

作　　者：陈佳义[1] 张琳[1] 

机构地区：[1]中山大学电子与通信工程系,06硕广东广州510275

出　　处：《中山大学研究生学刊（自然科学与医学版）》2007年第3期71-77,共7页Journal of the Graduates Sun YAT-SEN University(Natural Sciences.Medicine)

摘　　要：本文首先阐述平均熵在Turbo码迭代译码过程的性质,从中可以观察到译码算法的收敛程度与平均熵的变化趋势一致。然后,我们基于平均熵提出一种新的Turbo码迭代译码停止准则。仿真结果表明,对比现有的其它准则,这种新的准则能够在保持译码的误比特率性能和算法的计算复杂度的同时,有效地减少迭代次数。This paper firstly presents the Average-Entropy variation in the decoding process of turbo codes, by which we can have some insights into the iterative decoding. And secondly we propose a new stopping criterion based on the observations of the entropy variation in the process of turbo decoding. This new criterion shows good performance comparing with other well known criterions in that it requires less average number of iteration, while achieving similar BER and maintaining moderate computation complexity.

关 键 词：平均熵 TURBO迭代译码 迭代译码停止准则 

分 类 号：TN911.22[电子电信—通信与信息系统]