低复杂度FIR滤波器NR-SCSE优化设计  

NR-SCSE optimization design of low complexity FIR filter

作　　者：李成 郭涛[1] 石帅 畅彦祥 LI Cheng;GUO Tao;SHI Shuai;CHANG Yanxiang(Key Laboratory of the Ministry of Education of Instrument Science and Dynamic Testing,North University of China,Taiyuan 030051,China;China Aerospace Science and Technology Corporation(CASC),China Academy of Launch Vehicle Technology,Beijing 100000,China)

机构地区：[1]中北大学仪器科学与动态测试教育部重点实验室,山西太原030051 [2]中国航天科技集团有限公司中国运载火箭技术研究院,北京100000

出　　处：《电子设计工程》2022年第5期184-188,共5页Electronic Design Engineering

摘　　要：针对传统FIR滤波器设计中乘法单元占用大量资源的问题,基于非递归共子消除(NR-SCSE)算法,结合标准符号数(CSD)编码,提出了一种低复杂度的无乘法FIR滤波器优化设计方法。利用Matlab生成滤波器的原始系数集和输入信号源,用CSD和NR-SCSE对其进行变换,得到新的量化系数集。用Quartus算法语言实现了一个15阶低通滤波器,并在Modelsim中进行了仿真测试。结果表明,新的结合算法相比于传统算法使用的硬件资源更少,加法器数量减少了30%左右。In view of the fact that the multiplication unit in the traditional FIR filter design takes up a lot of resources,based on the Non Recursive Signed Common Sub Elimination(NR-SCSE) algorithm and combined with the Canonical Signed Digit(CSD) coding,a low complexity FIR filter optimization design without multiplication is proposed. Matlab is used to generate the original coefficient set of the filter and input signal source,and then CSD and NR-SCSE are used to transform it to get a new quantization coefficient set. A 15 order low-pass filter is implemented by Quartus algorithm language,and the simulation test is carried out in Modelsim. The results show that the new algorithm uses less hardware resources and reduces the number of adders by about 30% compared with the traditional algorithm.

关 键 词：FIR滤波器 非递归的带符号共同子式消除 CSD编码 乘法器 

分 类 号：TN911.72[电子电信—通信与信息系统]