I’ve been watching closely the recent integration of MLAT support into the piaware platform, I currently have 1 PiAware running at my property and whilst it reports MLAT information activity it never actually gets to the point of MLAT tracking via calculations.
07/06/2015 05:36:02 mlat(5560): Receiver connection: ready
07/06/2015 05:36:02 mlat(5560): Server connection: ready
07/06/2015 05:36:02 mlat(5560): Receiver: 49.5 msg/s received 1.0kB/s from receiver
07/06/2015 05:36:02 mlat(5560): Server: 0.0 kB/s from server 0.0kB/s TCP to server 0.3kB/s UDP to server
07/06/2015 05:36:02 mlat(5560): Aircraft: 16 known, 7 requested by server
07/06/2015 05:51:02 mlat(5560): Receiver connection: ready
07/06/2015 05:51:02 mlat(5560): Server connection: ready
07/06/2015 05:51:02 mlat(5560): Receiver: 42.7 msg/s received 0.9kB/s from receiver
07/06/2015 05:51:02 mlat(5560): Server: 0.0 kB/s from server 0.0kB/s TCP to server 0.2kB/s UDP to server
07/06/2015 05:51:02 mlat(5560): Aircraft: 17 known, 7 requested by server
07/06/2015 06:06:02 mlat(5560): Receiver connection: ready
07/06/2015 06:06:02 mlat(5560): Server connection: ready
07/06/2015 06:06:02 mlat(5560): Receiver: 39.9 msg/s received 0.8kB/s from receiver
07/06/2015 06:06:02 mlat(5560): Server: 0.0 kB/s from server 0.0kB/s TCP to server 0.3kB/s UDP to server
07/06/2015 06:06:02 mlat(5560): Aircraft: 15 known, 7 requested by server
Anyway, I assume the lack of MLAT calculations is due to lack of other MLAT enabled receivers seeing the same plane and providing the relevant tracking information, if this is the case I have the option to install another 2 PiAware feeders with external antenna at a friends place (site A) and the other at my sisters house (site B), the distances between the 3 locations are:
Home - Site A 12.4km
Home - Site B 10.4km
Site A - Site B 2.6km
They form a squashed triangle with the short side between site A and B, will this provide any benefit for MLAT tracking??? As I don’t have too many other sites to choose from.
Other PiAware feeders in the area either aren’t running the latest MLAT enabled version or have limited/unreliable feed data.
This is potentially useful but the errors will be quite large. The FA mlat parameters are also set up to require 4 receivers before generating results, so you’ll need to be able to overlap with at least one more receiver somewhere else.
That said, it’s probably worth trying if it’s not lots of hassle.
I see mlat results 20-30km north of Brisbane so hopefully you can get overlap with those receivers.
Yes the merge is currently happening before the data is sent to piaware. The setup here currently is
Home site: SBS3 → modesmixer → local piaware + fr24feed + basestation
Remote site: dump1090 → remote piaware + modesmixer@home site
Because of the locations of the sites I get a good picture locally as to all traffic, but my concern is that because the two sites are separated by a reasonable distance (~10-15km) the MLAT data being sent by my home site instance (piaware and fr24feed) will not be usable unless there is a way to filter out ‘relayed’ reports ?
If that is the case I will disable MLAT on my home site to avoid injecting ‘bad’ data.
This will not work for mlat. You must feed each piaware with data from only one receiver for mlat to work. With the setup above, the “home” piaware sees data from two receivers. Also, it will be seeing data with the wrong receiver location, which is very important for mlat to work correctly. Probably you’ll just never get clock sync on the home site.
I don’t know what modesmixer does with SBS3 timestamps, but they’re unlikely to be useful for mlat. (fa-)mlat-client does support SBS connections natively in a way that will work for mlat. but there’s no way to configure that currently via piaware.
I would turn off mlat on the “home” piaware but leave it running on “remote”.
You probably also want to think about putting the “home” piaware before the mixer step (I guess maybe you’ll need a second modesmixer to convert SBS → Beast format); currently you’ll be feeding positions from the remote site twice, which is not really harmful at the moment but will be distorting your stats.
Thanks - that confirms my suspicions… I’ll disable MLAT for now and have a think about how to reconfigure things so that my feeder apps are only sending the local receiver data. I do also run VRS which looks like it could act as the mixer and work here, but was trying to offload ‘core’ functions to the RPi so I can turn off the windows box and not affect uploading of data. Perhaps a second copy of mixer running on alternate loopback ports may be the answer…
I am amazed at the number of MLAT aircraft in the Fort Lauderdale (KFLL) area: almost 50% of 60 aircraft being tracked. Also very interesting are the sometimes jagged track lines showing the mathematics of more or less sites contributing to the MLAT calculated position. It is very important that all sites enter very accurate GPS coordinates and altitudes into their profiles. Kudos to all the developers and supporters of this project. Mere mortals like me really appreciate the resources on this forum. Joe K4AA
I just fixed a rather silly bug on the mlat server side, which in theory should mean that everyone should be seeing more mlat results now.
(My local receivers went from ~100/minute to ~180/minute)
I’ve never seen any ‘Results’ in the log nor any MLAT colored aircraft on the web display. Am I just too isolated with no one near to coordinate my MLAT data?
There aren’t any mlat results being generated near you unfortunately. Closest results are near Boston. I haven’t looked closely at why, but I’d guess it’s just a lack of nearby receivers with a common view of the sky.
There are enough people running the right version of PiAware near you, but they all have bad range
It would be good to get together with them to discuss antennas, but even if you PM them - there’s a fair chance they wont notice the message is waiting and the forum doesn’t mail them to let them know a message is waiting.
I was going to try and contact few but thought better of it. I’m going to try and set up a few more feeders at some different locations to see if that’ll help.
MLAT’s running about 40% of total (42 aircraft) right now. Lot’s of execjets in the South Florida area. Although there are commercial aircraft putting out Mode S only. Perhaps their Mode B is not turned on or is toast.
I installed Piaware 2.1-1 and Dump1090-mutability 1.15-dev late Friday local. Below are the graph of some early results. I have seen a ‘flip’ of my count of ‘with Positions’ vs ‘No Positions’ aircraft.
Before MLAT, ‘No Positions’ was the majority. After MLAT, ‘with Positions’ have become the majority. The spikes are caused by the periodic reception of TIS-B aircraft near KLAX.
Obj, looks like the server change did increase MLAT aircraft. From the graph, it appears that my ‘with positions’ aircraft increased today.
[quote=“mighty3507”]
I installed Piaware 2.1-1 and Dump1090-mutability 1.15-dev late Friday local. Below are the graph of some early results. I have seen a ‘flip’ of my count of ‘with Positions’ vs ‘No Positions’ aircraft.
Before MLAT, ‘No Positions’ was the majority. After MLAT, ‘with Positions’ have become the majority. The spikes are caused by the periodic reception of TIS-B aircraft near KLAX.
/quote]
I have noticed a huge increase on my local stats as well. Still not showing on FlightAware.
