采用近似max*运算的Log-MAP译码算法  

LOG-MAP DECODING ALGORITHM USING APPROXIMATE MAX*OPERATION

作　　者：孙增友[1] 李欢欢[1] 王蒙[2] 王景芹[2] 

机构地区：[1]东北电力大学信息工程学院,吉林吉林130012 [2]河北工业大学电气工程学院,天津300130

出　　处：《计算机应用与软件》2016年第3期255-258,共4页Computer Applications and Software

基　　金：国家自然科学基金项目(51077039)

摘　　要：为有效降低Turbo码译码的硬件存储消耗,提出一种基于近似max*运算的改进的Log-MAP算法。并通过设计合适的数字电路来找出一组数据中最大的两个值嵌入到其相关函数项中,有效实现了低复杂度的Turbo译码器的硬件结构。实验结果表明,所提出的结构比Constant Log-MAP算法结构平均简化了30%,达到了与Log-MAP几乎相同的误码率(BER)性能,降低译码的复杂度,便于实际工程应用。In order to effectively reduce the consumption of Turbo decoding hardware storage,in this paper we propose an improved LogMAP algorithm which is based on approximate max＊operation,and find out two maximum values in a set of data through designing an appropriate digital circuit and embed them into their correlated function terms,thus effectively implement the hardware architecture of Turbo decoder with low complexity. Experimental results show that the proposed architecture is simpler by 30% on average than the architecture of Constant Log-MAP,it reaches the BER performance almost the same as Log-MAP＇s,reduces the decoding complexity,and is convenient for practical engineering application.

关 键 词：TURBO码 LOG-MAP算法 CONSTANT LOG-MAP 复杂度 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术] TN911.22[电子电信—通信与信息系统]