Study of NLMS algorithm used in the cancellation of navigation interference in the HI observation

In the radio observation based on the large single antenna, the navigation signal interference is one of the most common radio frequency interference (RFI). In this paper, we use an auxiliary channel and normalized least mean square (NLMS) algorithm to reduce the interference of navigation signal L
2
from the observation of the red-shifted HI spectral line at L-band. Because the radio astronomy signal is extremely weak and the power of noise from the main antenna is more than that of the radio signal. The existence evaluation index cannot express the changes of each part of signals very well, we propose two indicators to evaluate the anti-interference performance more directly: the radio signal attenuation G
s
and the interference suppression G
I
. Furthermore, the fractional delay of navigation signal L
2
between the main and auxiliary antenna, which means a non-integral multiple of the sampling interval, can obviously influence the algorithm performance. Thus to remove the fractional delay we design a finite impulse response (FIR) fractional delay filter based on the prolate wave function, which can effectively suppress the sidelobes in the frequency domain caused by truncation. Ultimately, the performance of NLMS algorithm and the utility of the proposed indicators are validated by detailed emulation. And after the cancellation of the fractional delay, the performance of the NLMS algorithm is improved, and the radio signal attenuation is also decreased.