Comparison Of Identifying Algorithms For Spaceborne Radio Frequency Interference

The influence of radio-frequency interference on microwave remote sensing has become an increasingly serious problem.In order to better use the observed data,it is necessary to correctly identify and remove RFI contamination.To find out the applicability of various methods,using spectral difference method,double principal component analysis(DPCA)method,multi-channel regression analysis method and ratio method respectively recognize RFI signals in the Advanced Microwave Scanning Radiometer 2(AMSR-2)observations at 6.925GHz,7.3GHz and 10.65GHz channels and compared the monthly results of above methods.Research shows,in summer land,the spectrum difference method,DPCA method,and multi-channel regression analysis can effectively identify the interference of RFI signals in 6.925 GHz in the United States and 7.3 GHz in the European region.In the RFI identification of 10.65 GHz in Europe,the DPCA method is used in summer and winter.Both of them can be used.Spectral difference method and multi-channel regression analysis ean effectively identify RFI signals in surnmer.In winter,misjudgment can be caused by the effect of snow underlay surface.The ratio method can only identify strong RFI interference in England.When the C-and X-band observations have RFI interference at the same time,the ratio method will not recognize the RFI signal.In the ocean,for the RFI identification of the 7.3 GHz channel near the US and the 10.65 GHz channel near the European sea,the DPCA method,multi-channel regression analysis and ratio method all identify the RFI signals with the same distribution trend.However,the RFI signal identified by multi-channel regression analysis has a stronger intensity and a wider range.The law of ratios will result in misjudgment at the sea-continental boundary and in the vicinity of the field of view.The RFI index of the ratio method does not have intensity significance and can only be used to determine the existence of RFI signals.In short,the multi-channel regression analysis method and the double principal component analysis method have a wide range of applications,while the spectral difference method and the ratio method have their own drawbacks.