According to the HSP43168 chip manual, how exactly is this chip interpolation filter and decimation filter implemented? How to use FPGA design to achieve arbitrary 2~1000 times the sampling rate interpolation or decimation?