Doppler from ADS-B speed

[jehanazad]

Currently the doppler shift is determined by differentiating the positional changes from the ADS-B reported position.

When positional data is intermittent, this calculation fails and doppler is returned as 0. Currently, at any given time, most aircraft are marked as 0 doppler.

A solution is to calculate the doppler from the reported ADS-B speed and heading.

This will require some computation, as the shift will be relative to both the transmitter and receiver directions. Using ENU coordinates would be easier, but if this is too imprecise, ECEF coordinates could be used.

https://www.notion.so/adsb2dd-2a7e835d2f5480d8a1bbe75409706f7e?source=copy_link

[jonnyspicer]

@jehanazad I've implemented this, seems like it has mostly fixed aircraft showing as having 0 Doppler. There are still some remaining, but this appears to be due to issues with the ADSB feed, which sometimes has aircraft with 0 baro_rate and geom_rate.

LMK if it looks good to you, if so I'll close this ticket.

[jehanazad]

@jonnyspicer Yes, the 0 doppler ones are fine when there's not velocity data. It looks like that will happen some of the time.

On the delay-doppler map I now see nearly the green dots distributed across the map. They seem to be frozen in place though, while the orange detections come and go. Any idea what's happening there? On the ADSB feed I see some planes not moving much, but others are updating pretty frequently.