Final Report of BeJamDetect Project
Time evolution of PRNs in PVT and in track
SP FE
441
1.
3
a
A
NEE
An
bh
Alr
A)
=used for PT 5
= tracked, GPS L1-C/A
= tracked, GAL E1-B
7
4
09-hay
ı 2-May
15-May 1®-May
UTC time
El e.
21-May 24-May 27-May 30-May
2022
Figure 4-14: Number of GPS L1-C/A signals being tracked and used for PVT, May 2022
As can be observed in Figure 4-13 and Figure 4-14, the mean C/NO0 value and the number of tracked
satellites fall down significantly several time over the month. A detailed investigation of such
ootential interference events (see an example below in Figure 4-17) reveals that all they are of very
short duration. Typically only one measurement epoch is affected and the next measurement
coming 0.5 sec later contains nominal values. Such behaviour was, most probably, caused not by
external radio frequency interference but by interference from the internal electronics of the
measurement set-up.
Elgenvalues of pre-correlation array covarlance matrix
ir
L——
x
2
— |
=,
|
35
ln — — —mrnx_—_ _— .— BEN — .—
06-May 09-May 12-May 15-May 18-May 21-May 24-May 27-May 30-May
UTC time 2022
Figure 4-15: Eigenvalues of pre-correlation array covariance matrix, May 2022
Figure 4-15 shows the time evolution of all four eigenvalues of the array covariance matrix. The use
af the eigenvalues for interference detection has been already successfully tested in previous
measurement campaigns (see [14], for example). Because of very low received power of the GNSS
navigation signals in the interference-free case the received antenna signals before the de-
spreading process are dominated by the thermal noise originated from the active RF/IF electronic
cmponents of the receiver. The spread of the eigenvalues should be in this case small because all
aigenvalues span the same noise signal sub-space. In the case of jJamming, the antenna signal has
a strong Jamming signal contribution along with the thermal noise. This results in a wider spread
Fitle: Final Report
Version: 1.0
age: 49
