5G极化码的低复杂度SC译码方案  

Low complexity successive cancellation decoding scheme for 5G polar codes

作　　者：陈发堂[1] 唐成 刘一帆 CHEN Fatang;TANG Cheng;LIU Yifan(Communication and Information Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,P.R.China)

机构地区：[1]重庆邮电大学通信与信息工程学院

出　　处：《重庆邮电大学学报（自然科学版）》2019年第6期753-759,共7页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：国家科技重大专项(2012ZX03001024)~~

摘　　要：研究了对数似然比域(log likelihood ratio,LLR)中极化码的串行消除(successive cancellation,SC)译码算法的结点信息更新公式,即在信息更新时,需要对指数和对数等操作进行运算,复杂度较高,当码长N增大时,其信息更新复杂度增大.为了降低SC译码算法的复杂度,提出了量化和积算法和一种可用移位操作的等误差线性近似分段算法.运用量化和积以及等误差线性近似分段函数代替SC译码算法中的双曲函数运算,即仅需乘法和加法操作,降低了计算复杂度,易于硬件实现.仿真结果表明,提出的量化和积算法较SC译码算法的性能损失不超过0.1 dB.提出的等误差线性近似分段算法在增加少许复杂度情况下,性能更加逼近SC译码算法.当误码率达到10-4时,提出的算法比现有的分段线性近似算法有0.4 dB的性能增益.The node information update formula of successive cancellation(SC)decoding algorithm for polar code was studied in the log likelihood ratio(LLR).In the process of information updating the operation of exponent and logarithm is needed.The complexity of information updating is high.When the code length N increases the complexity of information updating increases.A quantized sum product algorithm and an equal error piece wise linear approximation algorithm with shift operations are proposed to reduce the complexity of the SC decoding algorithm.The proposed algorithms in which only multiplication and addition operations are required are easier to implement in hardware because the hyperbolic function op erations in the SC decoding algorithm are replaced.Simulation results show that the performance of the proposed quantized sum product algorithm is less than 0.1 dB compared with the SC decoding algorithm.The performance of the proposed equal error piece wise linear approximation algorithm is closer to the SC decoding algorithm with a little complexity.A gain of a bout 0.4 dB can be achieved in comparison with previous piece wise linear approximation algorithm when the bit error rate is 10-4.

关 键 词：极化码 串行消除 量化和积 等误差 线性近似 

分 类 号：TN929.5[电子电信—通信与信息系统]