IEEE Geoscience and Remote Sensing Magazine - June 2014 - 71

0

10

20

0

10

20

30

40

50

60

30
40
Time (s)

50

60

Single Algorithm
2 Algorithms Combined
3 Algorithms Combined
4 Algorithms Combined
5 Algorithms Combined
Uncalibrated Brightness
Temperature Spectrogram

Spatial
Temporal

0
20
0
40
0
60
0

Pulse
Algorithm
Flag

1.0
0.8
0.6
0.4
0.2
0

Vertical TB Deviation vs. Third Stokes Parameter
(Nadir Antenna)
Data Acquired 2008/04/14
600
Clean Data
500
400
Contaminated
300
200
100
0
-100
-200
-800 -600 -400 -200 0 200 400 600
Third Stokes Parameter [K]
Vertical TB Deviation vs. Fourth Stokes Parameter
600
Clean Data
500
400
Contaminated
300
200
100
0
-100
-200

Pulse Detection Algorithm

10
00
80
0
60
0
40
0
20
0

Uncalibrated
Brightness
Temperature

160
140
120
100
80
60

Fourth Stokes Parameter [K]

Stokes

128 Channel Spectrum

Statistical

120

1200

Spectral

110
100
60
176
50
174
40
172
170
20 Seconds
168
10
Normalized Kurtosis Spectrogram
166
Spectral Channel
164

Time (ms)

Uncalibrated TB

1400

1000
800
600
400
200
0 100 200 300 400 500 600 700 800

Normalized
Kurtosis

Frequency (MHz)
2.5
2
1.5
60
50
1
40
0.5
Seconds
176 174
20
172 170
10
168 166
164
Spectral Channel

Figure 3. A five dimensional Venn diagram [16] illustrating the different classes of RFI detection and mitigation algorithms and their
combinations. Two (red), three (blue), four (green) and five (grey) classes combinations are possible. The algorithms clockwise from top left
are: Pulse Detection (Temporal) [8], 3rd/4th Stokes RFI Detection (Polarimetric, courtesy of Skou et al. [10])), Spectrogram for CW signals
(Temporal + Spectral), Kurtosis (Temporal + Spectral + Statistical, courtesy of Ruf et al. [17]).

The algorithm is fairly easy to implement, but it is inherently biased and can only detect RFI above the noise floor
of the radiometer, potentially missing RFI coming from
continuous sources. Fig. 1 shows that it performs satisfactorily by picking up sources around the Europe, China
and the Middle East.
The SMOS mission employs a unique interferometric
technique for measuring brightness temperatures. This
results in an RFI detection approach that is a combination
of spatial, temporal (or angular) and polarimetric algorithms. The algorithm proposed by Skou et al. [10] looks
june 2014

ieee Geoscience and remote sensing magazine

at the third and fourth Stokes parameters of the incoming signal and any deviation away from zero (as seen
in Fig. 3 polarimetric). Another algorithm proposed by
Misra et al. [11] takes advantage of the smooth geophysical variation of brightness temperature with respect to
incidence angles, and detects deviations from a smooth
fit. The SMOS mission has been very active in identifying the interference sources at L-band, geo-locating them
and turning them off. This has led to more complicated
algorithms that not only detect RFI sources but attempt to
locate them as well [12].
71
Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - June 2014
