Research On Ionospheric Clutter Mitigation Method Based On Multi-resolution Analysis

High Frequency Surface Wave Radar(HFSWR) is a kind of instrument, which can provide over-the-horizon surveillance of maritime targets using HF waves(3MHz~30MHz) propagating along the earth curvature. Ionospheric clutter is one of the main interferences. Aiming at the clutter within the beam’s mainlobe as well as the targets, which the classical Generalized Sidelobe Canceller(GSC) cannot mitigate, this paper takes the advantage of MRA which is suitable for extracting local features of non-stationary signals and studies the ionospheric clutter mitigation method based on MRA.Firstly, the distribution character of ionospheric clutter in traditional time, frequency and time-frequency domain is analyzed. Then radar signals converted to MRA domain(1D wavelet, 2D-wavelet or ridgelet domain) is used for analyzing the distribution differences between targets and clutter.Secondly, two clutter mitigation methods are proposed according to the distribution feature. The one is clutter mitigation method based on ridgelet through time-frequency image processing, and the other one is clutter mitigation method based on wavelet through Range-Doppler image processing. This paper describes the principle and the processing of the two methods in detail and shows the preliminary results. Experimental results show that the former method can extract targets’ peaks almost submerged from the clutter and the signal-to-clutter ratio(SCR) can be improved about 2.65dB~12.06 d B. Then, the latter method can remove the large-scale clutter in Doppler spectrum, the SCR can be improved about 2.65 d B~12.06 d B.Finally, to take the advantages of the two methods, the clutter mitigation scheme based on MRA is proposed. The results are better than the ones using the two methods separately. Experimental results show that the effect of clutter mitigation varies with different clutter background. When the background is E layer clutter, the results is almost the best, the gain of SCR after the processing can reach about 25 d B. The higher the original SCR is, the higher SCR gain can be gotten after the MRA processing. By comparing with the spatial domain method, results verify that the MRA method can cover the shortage of spatial domain method. It can mitigate the mainlobe clutter in effectively. The detection results of batches data show that the detection performance can be improved by the clutter mitigation method proposed by this paper.