Abstract:To reduce the performance loss caused by fixed-point quantization of variable fractional delay filter coefficients using truncation methods and optimize the resource consumption of fractional delay filter engineering applications, this paper proposes a Farrow filter design method that combines the Minimax method and mixed integer linear programming(MILP). And, the multiplexing method of Farrow filters is introduced. Firstly, based on the Farrow filter structure and coefficient symmetry, the filter coefficient solution is characterized as an optimization problem with amplitude-frequency response peak error constraints and fractional delay error constraints and integer constraints are established on the Farrow filter coefficients. Secondly, based on the established constraints, the filter coefficients are solved. Then, resource consumption is analyzed based on the introduced Farrow filter reuse method. Finally, a 16-channel, 4-beam beamforming system experimental platform is built on the Xilinx VCU118 development board to validate the filter design method and multiplexing method proposed in this paper. The experimental results show that, under the same quantization bit width conditions, compared with traditional truncation quantization methods, the proposed method reduces the normalized root-mean-squared and group delay error of fractional delay filters. The Farrow filter bank has significantly optimized resource consumption after reuse. The digital beam pattern formed by the experimental platform meets expectations without any directional deviation. The design method and multiplexing method proposed in this paper can provide reference for the engineering application of Farrow filter in broadband multi-beamforming system.