glonass navigation satellites
aug14: Monitoring glonass for 3 days
dynamic spectra of over 3
data.. sat position power vs az,za
examples of glonass rfi in p2030 data
up to global navigation
The glonass satellites are the Russian global
|1246 + k*Fs
k=-7 to 13
1602 + n*.5625
n=-7.. + 6
n=satellite channel freq #
- standard accuracy code chip
- 2438.36 L2 cycles/chip
- .511 MHz
- pnCode. length=511 code chips, period=1 msec
- high accuracy code chip
- 243,836 L2 cycles/chip
- 5.11 MHz
Orbit: 3 planes of 8 satellites each
- inclination: 64.8deg
- each orbit spaced by 120 degrees in longitude (see RA
ascending node in tle).
- 19100 km altitude.
Aug14: Monitoring the glonass signal for 3
During tropical storm
Bertha (01-03Aug14) the telescope was locked in position.
lband wide data was taken to monitor the band during this time.
the setup and observation
- az: 257.5330 (encoder value)
- za dome: 8.4773
- ch : 8.8347
- lbw , linear polarization, full band (1150 to 1736 MHz0
- Acquisition time:
- Start:: 01Aug14: 21:11:42 (utc)
- end : 03Aug14: 11:25:00 (utc)
- 3570 seconds data
- 10 sec cal on
- 10 sec cal off
- (note some patterns were terminated prematurely because of
- mock spectrometer. 4 boards
- 8192 channels, 172.032 MHz bandwidth, polA, polB per board
- board cfr's = 1200,1350, 1500, 1650 MHz
- 1 second dumps
- N=1..4 for the 4 boards
- D=01.. 03 for the 3 days
- M=0..3 the 4 boards
- xxxxx file numbers..
- each file starts on a multiple of 100,
Dynamic images covering glonass band
dynamic spectra from each pattern (typically
3570 seconds) were made covering 1550 to 1620
Processing the data
- This includes gps L1 (1575), glonass, and the start of the
iridium band (1618).
The table below shows the dynamic spectra for each pattern:
- A bandpass correction for each image was made by:
- compute the median bandpass for the pattern
- do a robust fit (throwing out outliers) to 1550 to
- use a 2nd order polynomial and 5th order harmonic.
- Scale the flatten image to -3 + 6 sigma (full scale lut).
Notes on the dynamic spectra:
- The vertical scale is in 1 second steps
- the dashed vertical lines are the centers of glonass channels
- horizontal lines that cover the image are probably continuum
source drifting through the beam
- when a satellite gets close to the beam, the power spreads in
freq (we see more sidelobes).
Images of the satellite position and the power in each glonass rf
channel vs az,za were made:
- idl routines in x101/lbwmon
- 13oct14: split each dataset into separate subdir:
- 140801 - bertha
- 141013 - gonzalo
- ./pro - holds generic .pro
- doit.pro - copy in each subdir (140801,141013). Defines
parameters for each data set. These are used by the generic
.pro routines to get/save the data.
- for 01 - 03aug14 data (taken during storm)
- did 1 hour scan followed by 10 sec calOn, calOff.
- 460 files, 4files/scan (since 4 boards) --> 125 scans,
data/calOn,calOff per pattern so about 40 patterns
- procdat.pro - get
info on all files, hdrs, patterns taken. store in idl save
file for later use.
- fnmIAll[Nfiles]: fname list of all files taken
- sumI[Nfiles]: summary header from each file
- PatI[npat]: hold info on each pattern taken (pattern is
data scan and then cal)
- patId: scan number
- hasCal: 0/1
- nrowsD: number rows of data
- brdAr[4} info on each board
- cfr: Mhz of board
- Idat: index into fnmIAll for data file
- iCon: index into fnmIAll for cal on
- iCoff: indesc into fnmIAll for cal off
- fnmIAll, patI,fnmIall,sumI saved to patI.sav (idl
- to process data just restore patI.sav and go from
- procglonass.pro -
create idl save file for each pattern. Contains total power
for each sec for each glonass channel. also contains header
- for each pattern, select the 4th board (1650 band)
- keep 1570-1620 mhz
- make a dynamic spectra (optionally write to disc)
- compute the total power for each 1 second integration for
each glonass frequency channel
- remove 2nd order baseline using data outside the glonass
band (1593-1597) and (1608-1615)
- save the tpAr,sumI1,ldate to an idl save
- stored in mocksave/20140801_211142.sav
(yyyymmdd_hhmmss.sav) time stamp is start of pattern.
- inpsav.pro - Load
save files (generated by procglonass) into glar. one
entry for each second of data taken
- jd time stamp for this second
- tpch: total power for each satellite for this
second (from save files generated by procglonass)
- az: az position of sat for this second (to be
- za: za position of sat for this position. (tobe
- cmpsatpos.pro -
using time stamps from glar compute constellation sat
position for each second
- glAr from inpsav.pro has the timestamp
- calls satpass routine to compute constellation position at
1 sec intervals for 12 hour blocks.
- generates satAzAr[npntsTot,nsat], satZaAr[npntsTot,nsat] .
save to satazza.sav file
- npntsTot same as glar[npntsTot] so indices should be the
- plotsatpwr.pro -
plot sat positions vs az,za.. make image of power in glonass
freq chn vs az,za position on satellite..
- currenly only have 6 satellites mapped to freq channel.
need to find the rest.
- doit.pro : after
above run, doit.pro will inpsav,,restore satazza.sav and then
- Processing sequence:
- procdat.pro. generates file list -> patI.sav
- procglonass. compute power in glonass channels -->
mocksave/yymmdd_hhmmss.sav . 1 for each 1 hour pattern
- inpsav.pro : input save file sfrom procglonass.pro.create
glAr[npntsTot] .. to hold total power , time ,satpositio
- cmpsatpos: using timestamps from glAr. compute sat pos for
each sec, sat. Save to satazza.sav . (should put it in
glar.azpos,glar.zapos but don't)
- plotsapwr: plot sat pos vs az,za, make image of sat chan
- red triangle - the platform triangle
- red bar: the azimuth arm
- green dot. position of the dome
- dotted circles. every 4 or 10 degrees in za.
- Positions: different colors are different satellites
- I've only mapped 6 satellites to frequency channels.. so
that's all that i've plotted.
- za Range. with a 3 arcminute beam, it is difficult to see the
satellite pass thru the sidelobes when the entire image cover
+/- 90 deg. The smaller za ranges let you start to see the
on/off of the signal as it passes through the sidelobes
(although we may have been sampling too slowly at once per
- Probably need to work on increasing the dynamic range of the