(n,1,m)递归系统卷积码的盲识别  

Blind Recognition of (n,1,m) Recursive System Convolutional Code

作　　者：张立民[1] 刘杰[1] 钟兆根[1] 

机构地区：[1]海军航空工程学院电子信息工程系,山东烟台264001

出　　处：《电讯技术》2014年第9期1220-1225,共6页Telecommunication Engineering

基　　金：国家自然科学基金资助项目(61102167)~~

摘　　要：递归系统卷积码(RSC码)是Turbo码子编码器常采用的编码形式。针对(n,1,m)递归系统卷积码的盲识别问题,给出了完整的编码参数和生成多项式识别方法。首先,采用矩阵分析法先求得卷积码参数,避免后续识别生成多项式时因参数循环估计带来的额外计算量。然后,根据RSC码的特性推导出生成多项式识别模型,给出了模型的具体求解步骤,并对所采用的Walsh-Hadamard变换算法进行了介绍。最后,运用Matlab平台进行仿真验证。仿真结果表明,该方法容错性能明显优于常规方法,在误码率为10-2条件下成功识别概率能达到90%,对于Turbo码的进一步研究具有重要意义。Recursive system convolutional code（ RSC code） is the commonly used encoding form of Turbo sub-encoders. For the blind recognition problem of （n,1,m） RSC code,a method based on Walsh-Had-amard transform is proposed. Firstly,code parameters are estimated by analyzing matrix constructed by the received codeword. Then,the mathematical recognition model is derived according to RSC code’s charac-teristics,the specific implementation steps are provided and the principle of the method is introduced. The simulation results from Matlab show that the performance of the method is obviously more excellent in com-parison with other methods. The recognition probability reaches 90% when bit error rate（ BER） is as high as 10-2 ,which indicates the great significance of the proposed method for further research on Turbo code.

关 键 词：数字通信 信道编码 递归系统卷积码 盲识别 Walsh-Hadamard变换 

分 类 号：TN911.22[电子电信—通信与信息系统]