Design and FPGA Implementation of an Adaptive Narrowband Interference Suppression Filter

The sensitivity and spectral performance of radio telescopes are increasingly affected by deteriorating electromagnetic environments. Therefore, developing a real-time interference elimination algorithm is urgent for future radio observations. In this article, we developed an improved interference suppression filter algorithm named the improved least mean square (IMLMS). Unlike traditional algorithms, the IMLMS algorithm uses an exponential function varying with error to adjust the step size. The input signal is normalized according to its own average energy so that it can not only satisfy the slow change in the step size at an error near zero but also reduce the influence of the input signal size on the algorithm. The improvement degree of the target interference ratio and the interference suppression degree are used to evaluate the performance of the IMLMS filter. Compared with traditional algorithms, the IMLMS algorithm has a better target interference ratio and interference suppression. The IMLMS filter is implemented in a Xilinx UltraScale KU115 FPGA, and the design of the IMLMS filter is optimized to meet the real-time observation requirements of a radio telescope with a 6-m aperture antenna at the Chashan observatory. The timing performance reaches 187.5 MHz. Compared to existing design and implementation methods, the IMLMS adaptive filter has high timing performance and solves the problem of interference suppression that requires high timing performance in fields such as radar and solar radio. Th e IMLMS filter is applied to the Chashan radio telescope for real-time observation, which further verifies the practicability of the IMLMS algorithm.