Intro
Some examples of the rfi

• GPS
• Galileo
• Plot of gnss frequency allocations (.pdf)
• esa navipedia page with gnss signals (.html)
• 01apr09 Novatel description of gnss signals (.pdf)
  

Intro

    Global navigation satellite cause interference in the frequency range 1165 - 1300 Mhz. These satellite system include:

• GPS- US system
• glonass - russian system
• compass-2 - chinese system
• galileo - european system

Some examples of the rfi:

    Project a2030 (galfacts) took data on 24feb09 from 19:00 to 22:00 hrs (AST). Their setup was:

• Pdev spectrometer, on alfa multi beam reciever.
  
• 172.032 MHz band centered at 1300 MHz
• Full stokes spectra computed every 1 millisec, 4096 channels.
• Data decimated to .2 secs and then .6 seconds.
• 2 Gbyte data files generated every 65 seconds.
• Az = 180 degrees. Za moves 2 to 19 degs at  .025 degs/sec for a few hours.

images:

Dynamic spectra file_100 with strong GPS L2 and 1227.6 MHz. (.gif):
Dynamic spectra file 23 with glonass and compass (.gif):

Plots:

GPS L2

• The data used polB (which was stronger)
• The dynamic spectra have .2 second resolution.
• The horizontal scale increases in frequency right to left.
• 1330, 1350 is the FAA airport radar.
• 1290, 1270  is the remy radar
• 1232/1247, 1241/1256. punta salinas radar in modeA
• 1268.52 Compass satellite E6 signal (weak)
  
• 1248.2  Glonass L2. (1246 + 5*.4375) (weak)
• 1227.6  gps L2 strong.
• The telescope sat still until 21 seconds. It then started to move in za at .025 degrees/sec. This is why the phase of the signal it pretty constant until 21 seconds where it starts to vary.
  
• the solid white bar at 5-9 seconds is the cal being turned on.

• Page 1:
• top: average of 65 seconds. Black is polA, red is polB.
• Bottom: pk power 1227.6 +/- 2MHz vs row in file. The dashed vertical line at row 34 is the peak value in polB
• Page 2: 10  .6 second spectra around the polB peak value at row 34.
• Page 3: The az,za locations for the GPS constellation near 24feb09, 20:00:00 AST.
•  The satellite causing the rfi is GPS22 which transits at 20:11:51 AST.
• Page 4: satellite location and distance.
• Top:az.za of the telescope (black) and az,za  of GPS22 (red) during the 65 seconds of file 100.
  
• Bottom. The angular distance (great circle) between GPS22 and the arecibo beam.
  
• The dashed blue line is where the peak in polB occurred.
• The telescope sat at za 7.3 until row 30. It then started to move up in za at .025 degrees/secs . This caused the distance to increase.
• The tle's for these computations came from 18mar09.  This may cause the distances to be off.
  

GLONASS L2 and COMPASS.

• The data is PolA.
  
• The dynamic spectra have .2 second resolution.
• The horizontal scale increases in frequency right to left.
• 1330, 1350 is the FAA airport radar.
• 1290, 1270  is the remy radar
• 1232/1247, 1241/1256. punta salinas radar in modeA
• 1268.52 Compass satellite E6 signal
  
• 1248.2  Glonass L2. (1246 + 5*.4375)

• Page 1:
• Top: average spectra (65) seconds. Black is polA red is polB
• Bottom: The peak Power in 1248.2 +/- .5 MHz vs row in file (.6 secs/row).
  
• PolA is black. polB is red.
  
• The dashed vertial line is row 39 close to the peak value in polA
• Page 2: 10 .6 second spectra plotted around row 39 (polA peak)
• top polA, bottom PolB
• The vertical scale is db. The peak got to about 15 Db above the power in the 20 Khz Channel.
• You can see the signal out to about +/- 12 MHz from the center.
  
• Page 3: Az, za position plots for glonass constellation near 24feb09:20:00:00 AST.
• Cosmos 2413 (blue) was that satellite causing the rfi in file_23.
• The azimuths are those of the satellite (not the telescope).
• Page 4: Angular distance telescope position to COSMOS 2413 for file_23.
• Top: ra/dec  for telescope (black) and cosmos 2413 (red)
• 2nd: az,za for telescope (black) and  cosmos 2413 (red). The cosmos az's have been converted to telescope az (-180 deg).
• 3rd: angular distance telescope pointing to cosmos 2413.
• The dashed blue line is the row for the peak rfi in pola.
• The angle is great circle.
• The tle files used for the distance computation were from 18mar09. The angles may be a bit off.
  

• The compass (beidou-2) satellite is part of the global navigation satellite system for china. The first compass satellite  was launched on 14apr07.
• Page 1: Average spectra file_23 65 seconds.
• top: full 172 MHz band
  
• bottom: blowup around 1268.52 MHz.
• Page 2: 109 (.6 sec avg) spectra overplotted with offsets.
• You can see the satellite transmitting about half way up the page.
• The vertical black line marks 1265.52 MHz.
• The 1270 rfi is the remy radar.
• I couldn't fine any tle files for this satellite so i'm not sure how close it got to the beam.
  
• The dynamic spectra above shows some leakage near 1383 MHz and 28 seconds that looks like it is associated with the compass satellite transmission. 

SUMMARY:

• The 10 MHz wide rfi that has been seen around 1227.6 is GPS L2
• The 10 MHz wide rfi that has been in the 1243-1252 MHz band is Glonass L2.
• The 10 MHz wide rfi seen at 1268.52 is the Chinese compass satellite.
• When the satellite get within a degree or 2 of the beam, the rfi gets very strong. I need to measure the rfi strength vs distance from the beam with more recent data and tle files.
• These satellites are not always close to the beam. It would interesting to see if we could schedule  around the satellites when they pass overhead at AO. To do this we need to know:
  
• How close does the satellite have to be to cause problems.
• Is it also a function of az, za or just distance.
• How fast do the tle files pointing accuracy degrade. Or how far in the future can we use todays tle to predict where the satellite will be in the future for scheduling purposes.