基于稀疏约束的低复杂度可变分数时延滤波器  

Low-Complexity Design of Sparse-Constrained Variable Fractional Delay Filter

作　　者：王静雯 周文静 沈明威[1] 韩国栋[2] WANG Jingwen;ZHOU Wenjing;SHEN Mingwei;HAN Guodong(College of Computer and Information Engineering,Hohai University,Nanjing 211106,China;The 54th Reasearch Institute of CETC,Shijiazhuang 050081,China)

机构地区：[1]河海大学计算机与信息学院,南京211106 [2]中国电子科技集团公司第五十四研究所,石家庄050081

出　　处：《数据采集与处理》2024年第2期481-489,共9页Journal of Data Acquisition and Processing

基　　金：国家自然科学基金(62271190);江苏省自然科学基金(BK20221499)。

摘　　要：针对基于Farrow结构的可变分数时延(Variable fractional delay,VFD)滤波器需求解大量子滤波器系数这一关键问题,本文将稀疏约束理论引入滤波器的权系数优化中,研究具有稀疏系数的Farrow结构滤波器。在极大极小(Minimax)准则下,通过添加L1正则化约束项改进权系数优化模型,在系数(反)对称性基础上进一步增加系数的稀疏度。然后,采用交替方向乘子法(Alternating direction method of multipliers,ADMM)进行权系数迭代求解。仿真实验表明,本文提出的基于稀疏约束的VFD滤波器在保证高延迟精度的同时,乘法器和加法器分别减少了47.69%和58.60%,极大地降低了系统运算量以及复杂度。Since variable fractional delay(VFD)filter contains a large number of coefficients to be solved,this paper presents a study on sparse-constrained Farrow structure variable fractional delay filter.We add a L1 regularization constraint to further enhance the sparsity based on coefficient symmetry and optimize its frequency response to approximate a desired frequency response in the minimax error sense.In addition,the alternating direction method of multipliers(ADMM)algorithm is used to iteratively obtain the filter coefficients.Simulation experiments demonstrate that the proposed sparse-constrained VFD filter not only ensures high delay accuracy but also reduces the use of multipliers and adders by 47.69% and 58.60% respectively,thus lowering system computation and complexity greatly.

关 键 词：稀疏约束 可变分数时延滤波器 极大极小 交替方向乘子法 FARROW结构 

分 类 号：TN911.72[电子电信—通信与信息系统]