Implementing a scanning mode

A recurring problem when multiple cubesats are launched together is identifying which orbital elements belong to which object. Identifying the cubesats themselves is possible if they transmit at known frequencies, or if their data can be decoded, but linking them to a NORAD catalog number is usually only possibly once the satellites have drifted away from each other. The linking to orbital elements can be done much faster if Doppler curves can be directly compared to orbital elements. This is something I’ve done with my strf software.

The SatNOGS network can help here, as the global coverage of ground stations should allow for many Doppler curves to be obtained quickly, allowing a better comparison to orbital elements to link two line elements to satellites.

I’m suggesting that we implement a SatNOGS scanning mode that can be selected like a demodulator which will record spectrograms (waterfall data) with time and frequency information, that are uploaded to the network to allow comparison with two line elements. The scanning mode could be a simple gnuradio flowchart connecting the SDR to a waterfall generator but without the Doppler correction. It should cover enough bandwidth to fit the largest Doppler shifts, and should have enough time and frequency resolution to allow a good comparison with orbital elements (from experience 1 second integrations and a few 10 m/s velocity resolution work well).

Once new cubesats are launched, this scanning mode can be enabled by the administrators as a demodulator for the predicted transmitter frequencies of this satellite, and passes can be scheduled using the best guess orbital elements. If, after a few days, the satellite is correctly linked to the NORAD catalog this demodulator can be disabled.


Tagging this as related to this thread from @ka2upw :
Is it possible to schedule an observation without the doppler correction?

Yes, this sounds very similar to that thread.
Douglas KA2UPW