CoSMOS: Performance of kurtosis algorithm for radio frequency interference detection and mitigation

Abstract

Abstract-The performance of a previously developed algorithm for Radio Frequency Interference (RFI) detection and mitigation is experimentally evaluated. Results obtained from CoSMOS, an airborne campaign using a fully polarimetric L-band radiometer are analyzed for this purpose. Data is collected using two separate integration times, as a result of which sensitivity of the detection algorithm is measured. The impact of RFI on remotely sensed data over land and sea is also presented.

L-Band RFI as Experienced During Airborne Campaigns in Preparation for SMOS

“…In the fall of 2005, agricultural areas were covered in Australia. In the spring of 2006, the North Sea west of Norway was the target [13]. Finland was the target area in 2007 with campaigns over the ice-covered sea in winter and over the open sea in summer.…”

Detecting RFI using microwave radiometers

“…Often the data acquisition is constructed to take only a few samples per image pixel and consequently the removal of observed RFI disturbances is impossible without dropping the whole pixel and even adjacent ones. Presently it becomes more and more convenient to have a very high post-detection sampling rate or even to sample the signals prior detection at a sufficiently high rate, which allows to some extent the removal of RFI contaminations [3,4]. Nevertheless, there are cases where even such an approach fails due to the strength of the RFI signals or their presence in too much of the sampled data.…”

On the Detection and Quantification of RFI in L1a Signals Provided by SMOS

“…8, the content of these differential signals includes two contributions: the radiometric noise and the RFI contaminations. Indeed, for those points which have not been classified as outliers, 1096 in X polarization and 1089 in Y polarization, it is only radiometric noise, and it turns out that the mean value of δT X and δT Y is close to zero, the standard deviations are about 0.21 K and 0.19 K, and, in both cases, the kurtosis [12] is close to 3. Finally, it should be noted that the setting of σ is done by the procedure itself since the standard deviation of the measurements is known or can be easily estimated from the measurements themselves.…”

“…Indeed, the challenge is to detect contaminated samples with a non-Gaussian distribution since RFI is often generated by nonthermal mechanisms, whereas radiometric signals are composed only of the thermal brightness of the scene and the noise generated by the receiver. While the kurtosis has been found to be a reliable indicator of a non-Gaussian distribution of samples and, therefore, of the presence of RFI [12], plenty of normality tests can also be successfully used [13], provided that a sufficient temporal resolution of the signal is available, but this is not the case of the MIRAS instrument as explained at the beginning of this section.…”