Comparison of microprocessor-based and FPGA-based adaptive sample rate notch filters

It has been shown that adaptive sample rate filtering is an efficient and effective method of removing narrowband interference from broadband communication signals. A relationship between the adaptive sample rate gradient and the traditional gradient is presented. Two demonstration platforms to be compared are: an AMD 29200 microcontroller implementation and Xilinx XC4010 FPGA. Although intended for communication applications, the systems are demonstrated using an audio signal as the broadband signal and a sinusoidal interference is introduced and eliminated. Using a fixed notch filter with center frequency of π/2, the AMD 29200 microcontroller with 16 MHz clock cycle was shown to eliminate sinusoidal interference from 1.25 kHz to 2.5 kHz and a single XC4010 FPGA-based system shown to eliminate sinusoidal interference from 3.9 kHz to 7.8 kHz