I am close to a flight path for Logan Airport (BOS), with jets coming over within a mile or so at about 15,000 ft. I had noted a while back that I lose planes overhead occasionally, and a reply to my post suggested experimenting with gain. I’ve tuned it pretty well, but I’m still seeing this behaviour occasionally, and it appears to be with MLAT contacts only.
Ok, I think I can understand that since you want the correlating ground stations to be spread out to get sufficient TDOA measurements, and perhaps when overhead I lose a little signal strength and other stations may “take over” on the data feeds.
What I don’t get is that often the plane will just completely disappear for me. It’s tracking fine inbound, then maybe pauses and goes white and turned north on the Skyview* display, but never gets reacquired. Due to the actual proximity I’m sure I don’t completely stop receiving the aircraft signals, but perhaps the FlightAware receiving system has in effect counted me out and doesn’t bother to welcome me back when I have more proper data? It’s just that those planes never show up on my Skyview server again, even though they are on the FlightAware tracking site and should certainly be visible to me.
Can any system developers involved in the MLAT processing chime in as to why this is happening?
Thanks,
Erik
*BTW when I first set up my station I had to laugh that the web display component is called Skyview, as my feeder station is named Skyview since I live on Skyview Dr…
There are more than enough receivers for MLAT of aircraft higher up, especially in areas with many feeders.
A post about when you actually get MLAT positions:
Basically with many receivers around you can’t expect to get full MLAT tracking.
Which is really a shame -.-
Another possibility is that you’re actually seeing an area of bad mlat coverage (for somewhat complicated reasons, aircraft directly overhead are a particularly bad case for the mlat math); you could check that by seeing if there is good mlat tracking for the flights you see go missing on the main FlightAware website at the problem time. But given that you’re in a densely covered area that seems unlikely.
Ok, but I guess it would be nice if I would get the positions as long as I’m sending data that matches that from other stations, even if I’m not being used for the calculations. I really like to track flights as a hobby and my local server is real-time, whereas the aggregate data on the FlightAware site is delayed…
Nope, every time I see it and go to check the main site there is no problem there, so I’m just not getting the data fed back to my feeder for those flights after I have an outage such as being directly overhead. Next time I’ll run view1090 with --show-only in order to verify I’m still seeing packets from that flight.
You can check that in SkyView as well.
In the table there is a column named “age”.
That’s how many seconds ago you received something from that aircraft.
So I just happened to see this occur again. The flight was already past my feed location, then its progression on the map stopped, but I was clearly still receiving data from the plane, and continued to do so for several minutes. The map table showed the altitude changing, but the location was not in the table, and the distance from my location was not changing. The flight was being tracked on the FA site.
I ran view1090 --show-only and captured quite a large number of packets, and the signal was quite good (better than -3.0 for a while).
As i wrote that’s expected with many receivers in the area.
Your site was used to get at least one location of that plane.
After that you receive the position for 3 minutes.
If the site is not used again you’ll not get the location anymore.
Yeah, I guess I’m maybe asking the FA folks to enhance it as a courtesy. If I’m sending data for a non-ADSB flight, periodically send the position information derived via MLAT to me anyway, even if I’m not acting as a site used in the calculations. Or make it configurable for those sites that want to serve up local data and not just act as a feeder only.
