A GPS navigator connected to the navigational system and transponder of a plane must correct for such errors. A plane will hear more than 4 satellites all the time and some cross checking can be done.But what if the GPS system is wrong as in January with more than 10 us error for one day ?
You forgot the downlink from the plane (transponder) to the ground receiver. These messages have quite strong CRC checking and it is very unlikely that a faulty message be declared "good" if all available checks are done by the downlink receiver. This downlink is affected by ground reflections etc. I assume that downlink receivers used by ATC will do full CRC checks before passing the message further. However, some hobby receivers might not do full CRC checks, before passing the data as "good".
This has been seen with older planes without integrated GPS but relying on less accurate navigation system such as inertial navigation system. A few kilometer accuracy is sufficient for finding the ILS beam, but the transponder still sends out the inertial system coordinates.
Kinda sounds like a pseudolite might be involved: They're used to provide more accurate GPS positioning near airports so that GPS can be used for blind landing. However, I'm guessing as I have no experience with these.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
ADS-B is a secondary use of the transponder downlink (1090 MHz). The primary function is to respond to SSR (Secondary Surveillance Radar) uplink queries on 1030 MHz.
The ATC system determines primarily the position of the transponder by classical radar means (direction and time to response). There is also a system of measuring signal delays to multiple receivers and using triangulation (multilateration). Neither of these systems depend on the ADS-B data.
...and 978 MHz. I used to work on the software for the "official" ground radios that feed data to the FAA. I don't work for that company anymore. Opinions in this post are mine alone and do not represent the official positions of anyone else, nor any company.
The first cut at ADS-B involved the aircraft transmitting on 978 MHz, with a so-called "Universal Access Transciever" or "UAT". The box on the aircraft gets data from the aircraft's GPS and other instruments, and then transmits a single, relatively large, packet. This single packet has pretty much everything you'd want to know in it; the receiving end can decode just one packet and have a pretty complete idea of what the aircraft is doing. There are some consistency checks that are applied between successive packets from the same aircraft, but in general, it works well.
The FAA was ready to mandate UAT for everybody. However, a few years prior, the FAA had already mandated upgraded capabilities for the
1090 MHz radios that were already installed in larger aircraft (an outgrowth of the original transponders that transmitted a 4-digit code after being illuminated by the sweep of ground radar on 1030 MHz). The airlines had *just* gotten done spending money on that, and didn't want to buy more new radios for ADS-B, so they told the FAA to make ADS-B work on 1090 MHz.
The problem is that the 1090 MHz protocol was designed in the 1960s, when bits cost twelve million dollars apiece, so the goal was to use as few of them as possible. Newer extensions to the protocol let you use a few more bits than the original, but still not many. As an example, the full lat/lon of the aircraft don't get transmitted all the time; only the LSBs are, to save space. It takes roughly 3 to 5 1090 MHz packets to get one full ADS-B report. The receiver has to hold several packets in memory, and implement age-out rules for different fields in every packet, in order to be able to provide a report. For some fields, it's OK to report "old" data if you have it; for others, you either report very recent data or nothing. The rule set is complicated and not always consistent.
If we made a change to the receiver's UAT code, we could usually get it tested and signed off on the first pass. Changing anything on the
1090 code was a long slog, that sometimes involved pushing questions all the way back up to the FAA about how it was supposed to work.
There is also a *lot* of traffic on 1090 MHz. We had an outside antenna in the lab, and in sunny Kansas City, the 1090 "receive" LED on the radio would flicker off and on as airplanes went by. I heard a couple of reports from the field (mostly the east coast of the US) that the
1090 LED would come on solid in an actual installation, but I didn't believe it until I saw it myself. As soon as the installer hooked up the antenna cable, that LED came on and never went out again.
The way it is shaking out in practice, airliners have 1090 MHz, and most smaller aircraft have UAT/978 MHz. The ground radios take what they hear on 1090 and retransmit it on 978, and vice versa, so the aircraft all know about each other. There is also a facility to tell ADS-B- equipped aircraft about other aircraft that are only being picked up by radar.
Some of the ground radios are located relatively close to existing air-traffic control radars. These radios have a 1030 MHz receiver in them as well; it picks up the radar beam sweeping by, so the ADS-B radio can temporarily shut off its 1090 MHz transmitter, so as not to "blind" the radar.
The ground radios take the data they pick up over the air and forward it to a company (not the FAA) over phone lines. That company then consolidates the data from all the ground stations, and provides data feeds to the FAA, or to whoever else wants to pay for one (it's somewhat privatized).
As has been hinted at, the ground radios "trust" whatever the airplanes are saying. It's not impossible to spoof an ADS-B transmission. Part of the answer to that is that the ADS-B data can be correlated with radar data. Another part of it is for the mothership to compare the time-of-arrival of the ADS-B packets at several ground radios, and decide whether the triangulated aircraft position matched the self- reported aicraft position (this is called "multilateration").
The good reasons for ADS-B are to take advantage of better information that is available from GPS (as opposed to radar), and to provide better air-traffic control coverage in areas where it is difficult to get good radar coverage. The FAA's initial test was in Alaska. Many of the early production deployments were around the Gulf coast, to provide visibility for all the helicopter flights out to offshore oil rigs.
The real reason for ADS-B is to implement toll roads in the sky. Australia went with ADS-B for this reason. Every time you fly an airplane there, you get a few dollars added to your air traffic control bill; at the end of the month you have to pay up. The commercial airlines don't care; they just pass the cost along. General aviation doesn't like it. This hasn't yet been done in the US, but Congress is just starting to make noises about it.
Another subtext, which has been running since at least the late 1980s, is the argument of "now that we have GPS for airplanes, we can turn off and throw away all those radars, VORs, DMEs, ILSs, etc, and save tons of money on the light bill and maintenance", versus "it's a lot harder to spoof a radar return than a GPS satellite or an ADS-B transponder, radar still has a place". Currently, I think the status is still a little more on the "keep the radars too" side, but as time goes by, it may shift the other way.
That was probably far more than you wanted to know about it. :D
Again, I don't work in this field anymore. Opinions in this post are mine alone and do not represent the official positions of anyone else, nor any company.
There is another reason to keep the response packet short: The shorter the packet is, the smaller is the probability of two SSR responses running on top of one another, called garbling.
Initially, there were 4 different downlinks proposed for ADS-B. The most common is the SSR response (mode S extended squitter) at 1090 MHz.
The situation with 978 MHz and 1090 MHz is valid for the USA, but in the other areas ot the globe, the FAA UAT system is seldom used.
Thanks Matt, your post helped me to solve several mysteries.
Looking at the 978 MHz UAT message format, it really looks as good as
about the 1090 system with short message segments.
It seems that only a few parameters needed is transmitted in a single frame and multiple frames are required to get the total parameters of the plane.
It also explains, which sometimes you get a normal east-west path, then one sample about 100 km south and then the next sample close to the original path. Clearly, the 1090 MHz receiving software has guessed the latitude off by 1 degree which is 60 nm or about 110 km.
And the hobbyist receivers will mix up the situation.
No doubt, some (unreliable) heuristics is needed.
At least the 978 MHz frame contains both uplink and downlink data during the 1000 ms cycle, so the uplink data would keep the LED constantly lit.
If actual digital T1/E1 TDMA telephone lines (1.5/2 Mbit/s) are used, the propagation delay uncertainty is only abut 1 us, where as any packet switching system used e.g.for internet and hobbyist receivers would have much worse timing uncertainty.
In high reliability systems using double and triple redundancy with same type of equipment make sense, but if higher reliability is required, subsystems implemented with different technologies should be used.
This is true. If I remember correctly (it's been 3+ years), one of the tests we had to run on the 1090 MHz receivers involved deliberately creating some garbling, to make sure the receivers handled it properly.
In the "real world", garbling is apparently pretty common.
Yes, I should have said that my comments mostly applied to ADS-B in the USA.
Australia may have been the first country to put ADS-B in production. Western Europe was also implementing it as well. The company I worked for used the same receiver hardware in all three markets, but they had somewhat different software loads.
I don't know if there are many aircraft using the 978 MHz UAT frequency in Europe. UAT transmissions are synchronized to GPS seconds; IIRC the ground talks during the first 33% or so of the second, and the planes talk during the remaining 67%. If you hear data on 978 MHz that follows that pattern, then that's what you have.
I don't remember if the TV tuner stick implementation existed when I was still working at that company. I think somebody had a load for the GNUradio SDR platform that would do ADS-B.
If all you have is the hex code, and you think it might be a US- registered aircraft, you can grab the FAA registration database from
formatting link
and look for the "Mode S code". Note that sometimes the FAA quotes the Mode S code in octal, rather than hex. I don't know if there is an equivalent lookup for Europe-based aircraft.
If you know the registration number (N number), the FAA has a looker- upper for that, which also gives you the Mode S code. You can't go the other way, though, at least not via the FAA's site.
The Mode S code is in most of the 1090 MHz packets. The flight number is only in one or two types of 1090 MHz packet. If you only ever get a couple of packets for a particular aicraft, you might only know the Mode S code.
That's probably true, for hobbyist-type software. There are rules for unambiguously decoding the latitude and longitude, but they are fairly complex.
The UAT receiver was smart enough to only turn on its "receive" LED if it got downlink data. So it would normally blink on and off.
Land lines do have known latency, but the mothership didn't depend on the times the packets showed up on the phone line. The receivers had a high-resolution clock, locked to the GPS second, that was used to timestamp the packets received over the air.
There were latency requirements for the receivers that we had to test for - a maximum amount of time from "RF in" to "packet out".
On a sunny day (Mon, 8 Feb 2016 09:26:02 -0000 (UTC)) it happened snipped-for-privacy@att.net wrote in :
I will use my rtl-sdr spectrum analyzer to look at 978 MHz here in Europe.
formatting link
I really tried gnu radio and could not get it working on any PC or laptop here...
The 'dump1090' is from 2012 according to the author, and just a few lines of code (on top of rtl-sdr). Gnu radio is a big monster, I am sure it is fun for some, putting blocks together.
-rwxr-xr-x 1 root root 100093 Jan 21 2013 /usr/local/bin/dump1090*
so about 100 kB
I just downloaded that database, the S codes codes seem to be in hex. Now I can just do 'grep -i 4ca803 MASTER.txt' and see that code is likely not US registered :-)
Thanks for pointing to that database, 4sure I will lookup planes now every now and then.
I do notice that the location is only send occasionally, takes some time for dump1090 to display it. The range is really good on my indoors passive DVB-T antenna, say a hundred miles. Altitude helps :-)
On a sunny day (Mon, 8 Feb 2016 09:26:02 -0000 (UTC)) it happened snipped-for-privacy@att.net wrote in :
Some googling found that US hex codes start with 'A' (that database confirms that) as there seems to be no international standard, some people have researched it,
formatting link
so 4ca803 is UK (starts with '4', and RYR5VJ Ryanair I am sure ,so that figures for UK.
That *might* be it. We always just said "978 MHz" but I don't know what the precise frequency is.
If you take your spectrum analyzer to an airport that gets a lot of business jet and single-engine plane traffic, instead of airliners, and you see more activity on that frequency, that's a good clue. Try not to get arrested for standing at the end of the runway with a laptop. :D
I remember for sure that the 1090 MHz and 978 MHz signals used different modulation. 1090 might have been just keyed CW, or something relatively simple... I recall that the effective transmitter duty cycle was pretty low, like 15% +/-.
I honestly don't remember what the modulation for 978 MHz is. It is probably defined in one of the RTCA specs, but those are secret. There might be an ICAO or DFS or Eurocontrol spec that has it.
On a sunny day (Wed, 10 Feb 2016 05:55:25 -0000 (UTC)) it happened snipped-for-privacy@att.net wrote in :
hehe I am close to Leeuwarden mil airport: https://www.google.nl/maps/@53.2253401,5.753738,15z?hl=en
On a normal day there are many people at the start / end runway with radios, cameras, scanners. There even is a parking place there, been there many times: https://www.google.nl/maps/@53.2341612,5.7742247,607m/data=!3m1!1e3?hl=en
Some camera work:
formatting link
No arrest, have not brought this one there yet:
formatting link
Yes maybe I will do some googling on that. Mostly curiosity.
The 1090 MHz signal has two modulations: on-off keying for the classical SSR transponder response and BPSK for the Mode S data packets, including the ADS-B Extended Squitter data.
Interesting. I looked around the area on Google Maps. The image resolution for houses and buildings outside of the airport is as good as anywhere else in the world - probably less than 1 meter. I can see the hitches on the trailers at Titan aanhangwagens on Bredyk/N357, for example.
However, if you scroll over the runways and buildings at the airport, the image resolution is... worse. Hmm. (It's the same for google.nl and google.com, for what it's worth.)
I looked at the friendly neighborhood US Air Force base on Google Maps, and it seems to have the same image resolution as the surrounding area. Probably made the little purple light blink, too.
It looks like Reinhardt gave a good link for that.
On a sunny day (Mon, 15 Feb 2016 10:54:04 -0000 (UTC)) it happened snipped-for-privacy@att.net wrote in :
Yes, OK, well it is a rather open place, in the weekend they fly gliders there, also there is a model airplane club that can use it IIRC. There have been open days for the public, but maybe they did not want too many details shown, it is a NAVO base after all. But no kidding, one nuke does not have to be that precisely on target...
Interesting place in a way. I did see one emergency landing of a F16. An other F16 crashed just a few miles from where I lived in the water, pilot dead.
But.. that is the way it is, a 747 crashed into a flat in Amsterdam just blocks away from where I used to live. I once worked at Schiphol airport... Lots of stuff I cannot talk about. And that is civilian.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.