Structure design of fine-grained two-parallel Systolic FxLMS filter based on FPGA
Jun Yuan,Xiangsheng Meng,Qiang Zhao,Jun Li,Wei Wang,Guanyu Wang,Xiaoying Song,Qin Li,Jun Yuan,Xiangsheng Meng,Qiang Zhao,Jun Li,Wei Wang,Guanyu Wang,Xiaoying Song,Qin Li
DOI: https://doi.org/10.1016/j.micpro.2022.104604
IF: 3.503
2022-09-01
Microprocessors and Microsystems
Abstract:In order to improve the feasibility of feedforward FxLMS algorithm in active noise reduction headphones and improve the convergence, throughput and power consumption of FxLMS filter, this paper proposes a fine-grained two-parallel Systolic FxLMS filter structure design scheme based on FPGA. Two-parallel fast FIR filters are used to improve the throughput of the system. Multi-operand addition, Multiplication-accumulation unit and Fusion multiplication-addition unit are introduced to compress the number of operations, so as to reduce the critical path delay. Experimental results show that the throughput of this structure is twice as high as that of the traditional structure at the same frequency, and the convergence is close to that of Systolic FxLMS algorithm; compared with the existing optimal structure, the clock speed of the proposed 8-tap filter is increased by 52.17%, and the power consumption is reduced by 9.28%.
computer science, theory & methods,engineering, electrical & electronic, hardware & architecture