We have observed some random interference on downlink of C band carriers in India region. Duration of the unwanted signal is very narrowband, just like a spike of 1 or 2 second. But the amplitude is quite high & signal gets distorted. Also the frequency of the unwanted signal changes very often, not always at the same spot. Can anyone help us to sort this out?    

Do you have a plot you can share? In max hold or a short video clip showing the spikes?

Do you see it at a single station or at multiple stations?

Yes, I am sharing the downlink from an spectrum analyzer, you can see the normal , max hold & limit wave form. You can see some large spike at certain frequencies. To share the nature of these spikes are - a) they are short lived (duration of each spike < 500 ms) b) they  don't appear at the same frequency all the time c) when they start to occur the interference continues at same spot for 20/30 minutes. d) We received the unwanted signal on both polarization at same frequency (Horizontal & vertical) e) It is on insat C band (4.5 - 4.8 GHz) f) it is found in entire beam. Checked at several locations & found everywhere.   

---

https://www.satsig.net/yabbfiles/Attachments/compare_Vertical___horizontal.jpg
Regarding the two different polarisation spectrum plots above, in the left plot the burst of blue spikes between the pair of wanted yellow carriers suggests testing in preparation to set up a third yellow carrier in the middle. Maybe the cross-pol spikes (right plot) are because the linear polarisation feed has not yet been adjusted in rotation.
The blue, near central, triangle marker and central spike show good polarisation alignment and may be something quite different. Is anyone testing a new antenna for sidelobe patterns?  What uplink/downlink frequencies does the NOC allocate for such testing.

https://www.satsig.net/yabbfiles/Attachments/spectrum_analyzer.jpg
Regarding the spectrum analyser plot above, I can see an extra spike affecting the carrier in transponder to the left of the one with 6 smaller carriers.  The display spectrum appears to be inverted.

It is possibly helpful that the interference stays for 20/30 minutes as this will allow you time to home in on it and investigate further. Pulse like displays at different frequncies may build up in max hold to reveal a wider bandwidth transmission. If the interference is CW or a slow sweep then testing is a probable source. A CW signal varying might be a sidelobe test. If you can isolate a modulated TDMA carrier slot you can use ZERO span and observe burst envelope magnitude and shape.

If you have a limited number of antennas transmitting in 6.725-7.025 it might be worth asking if any were having work done around the time of the interferences.

Do you have more than one satellite operating 6.725-7.025 uplink ?

I hope Nimrod can add more.

The spikes are across quite a wide range and there are several of them. Indeed it is very curious.

I am sorry I only have more questions, but they might lead to something:

How many will be in the air at the same time? Only one or few?

It is hard to see at this span, how narrow are they? Like a CW or they carry some modulated data?

Do you know if they can be seen over other satellites? can they been seen free in the air? For example, connecting the spectrum to an LNB and a feedhorn but without a reflector - just as you would check for terrestrial signals for example?

Thanks for your analysis.
we have one satellite operating at this band, there are only 2 TX stations & no other stations are doing any test with this satellite so far we know.
HI Imrod, thanks first of all. i couldn't understand this question. "How many will be in the air at the same time? Only one or few?" can u pls clarify
I will check the span of the signal. The source is not terrestrial , that is pretty sure. To check if they can be seen o the other satellites, which satellites shall we choose - those close to our orbital location or those which has beam over our coverage region?
Can you guess the source of this unwanted carrier? Is it from another GEO or LEO or GEO earth station or LEO earth station - how can we narrow down the search for source?

It is very strange how most of the interference spikes do not appear to be affecting any of your wanted signals. Maybe just normal testing using specific unused frequencies to avoid causing problems?

Find an interferer, centre on it, set the RBW the same (or a somewhat wider to allow for frequency drift) as the interferer and go to ZERO SPAN, manual single sweep, and try sweep times of 1 mS, 300mS, 1 S, 1 minute, 1 hour. Keep pressing single sweep until you catch something interesting.

How can we get a list of other satellites operating at the same frequency band? or can there be any other applications who are using the same spectrum, 6.725 - 7.025 uplink or 4.5 - 4.8 downlink

I would aim a 1.8m dish and feed plus 4.5-4.8 GHz LNB at each of the INSAT satellites at 73.9, 82 and 85.5E. Compare spectrums and both polarisations.

For local sources (radar, sparks, 5G etc) put a 4.5-4.8 GHz LNB and feed together and hold in your hand while pointing around.