Decoding DME aircraft radio navigation system with the LimeSDR | Hacker News Reader

Decoding DME aircraft radio navigation system with the LimeSDR

destevez.net

117 pointswolframio38 comments

VORs are pretty interesting. At first i thought every plane had a small antenna array on board to check the direction to the VOR.

But the direction is just calculated by the phase shift between an omnidirectional and a directional signal. So it can be implemented very cheaply on every plane.

ADF uses radio direction finding techniques. It's not very accurate and is all but deprecated except for some corner cases where the legislation hasn't yet allowed for overlay approaches (replacing a real radio beacon with its true position via GPS / IRS).

There are some convincing reasons for being cautious on general on that front. Accidents have happened because some beacons have been substituted for the wrong location or even things like slant range being different from true range. In general ADF is so inaccurate that its protected areas are massive.

There's also a very real chance that commercial aviation may find itself operating in a GPS-denied environment, at least in various edge cases. For example, combat zones often employ GPS spoofing and jamming, and because radio waves don't exactly respect borders, this can sometimes affect civilian equipment outside of the combat zone. This has happened at least near Israel, Iraq, and Ukraine in the last few years. In other cases, truckers or rideshare drivers that want to spoof their company trackers end up parking too close to an airport and impacting planes on an RNAV approach.

> There's also a very real chance that commercial aviation may find itself operating in a GPS-denied environment, at least in various edge cases.

Yea, it's kind of terrifying that we are slowly putting all of our eggs in the GPS basket. I love GPS but when lives are at stake, you need a redundant backup navigation system that is robustly deployed and reliably works.

The VOR Minimum Operation Network[1] in the US is basically supposed to be that. They're decommissioning a lot of the VORs but at least guaranteeing that you'll be 100NM away from a working VOR and an airport with an approach that can be accomplished with VORs for the initial fixes.

Still definitely feels like putting a lot of reliance on GPS but at least there's a backup for the worst case.

[1] https://www.faa.gov/about/office_org/headquarters_offices/at...

There's also a DME Minimum Operational Network, for airliners that can use DME-DME RNAV. (That's too expensive for smaller aircraft to install, though.)

It's too bad that DME/DME RNAV isn't more widely available. The only real reason it's so expensive is that there isn't much demand for it since GPS (usually) works fine. Electronics-wise, it's not much more complicated than a transponder. Unlike a GPS, it does have to transmit, so it will always be somewhat more expensive than GPS.

The other problem is that there's a limit to how many aircraft a DME station can serve at a time (about 100), but I believe that could be greatly expanded if aircraft weren't pinging the DME so often. A position fix every second is generally fine, and it could be even more infrequent if you have a cheap inertial system to fuse with it that can fill in the track for a few seconds between pings.

One thing I have yet to understand is why DME-DME is preferred over VOR-VOR. Because the latter can support unlimited aircraft, unlike DME.

Probably the required accuracy. VOR is on the order of a degree for accuracy. DME is around 0.1nm. So if you’re 50nm from the VOR, then you may have a position fix error of 0.87nm across the radial, if I did my math right.

That's a good point. But if you're 50nm from a DME, I think you're unlikely to be able to get a lock in practice.

For altitudes above 12,900 ft AGL, the official service volume for a DME is 100-130nm.

Below that it's considered "line of sight"... and some quick math shows that you'd be able to get line of sight >50 nm for all altitudes above 1700 ft AGL (which is very low).

Source: https://www.faa.gov/air_traffic/publications/atpubs/aim_html...

Yes but only to about 100 aircraft at a time, favouring stronger (and therefore closer) signals: https://en.wikipedia.org/wiki/Distance_measuring_equipment#S...

If you're 100nm away, chances are there are more than 100 aircraft nearer to the DME than you from at least one of the two required DMEs. Especially if GPS has failed and many aircraft are trying to use backup DME-DME. Unless you're in a very sparse area.

> you need a redundant backup navigation system that is robustly deployed and reliably works.

Could existing cellular network base stations be this ground based backup navigation system?

https://people.engineering.osu.edu/sites/default/files/2022-...

Sure it could be done but good luck convincing cell phone vendors to adopt liability for aircraft safety

No. After about 5000 ft AGL (give or take) you can't pick up cell tower signals at all, since the antennas are directional and pointed towards the ground.

This is a deliberate design decision, because even a low-altitude aircraft would have hundreds of cell towers in sight and would overwhelm the network when handsets tried to register on all of them.

But also: Pilots like being able to have guarantees about system accuracy. We get notices anytime even a single GPS satellite is out of service (even though there are 31 of them), and have software tools in the aircraft to predict if there will be any signal degradation along our route (RAIM). I can't imagine having anything near that level of guarenteed safety with an ad-hoc system like described.

I agree. Starlink could potentially be used as an alternative to GPS (and other similar constellations). But it will probably be a long time, if ever, before it's certified for civil aviation navigation. And it's also vulnerable to jamming.

https://www.technologyreview.com/2022/10/21/1062001/spacex-s...

Starlink and GPS always face denial of service from exterior forces, or simply not being able to launch replacements up into orbit in a timely manner.

We should ALWAYS have full coverage from ground based navigation systems, and the pilots should be required to use them on a regular basis.

It's currently happening in the Eastern DRC too, affecting not only scheduled commercial traffic but humanitarian flights as well.

I'm wondering if there's some doppler-shift assessment possible to detect (and possibly reject) spoofing/jamming? A fixed ground-based spoofer/jammer will be too consistent in its doppler-shift, unless it varies its frequency slightly, but that would only work well against specific targets, not broad areas.

> Radio waves experience Doppler shifts the same as sound waves do as objects move. The Doppler shifts for the real satellites will all be different as the object moves either towards or away from them depending on their position in the sky. However, the Doppler shift caused by the object’s motion due to a spoofer is the same for all the satellites signals because they are all arriving from the same direction. This uniformity of Doppler is another indication of spoofing. Again, only the most sophisticated spoofer can account for an object’s motion to adjust its Doppler shift for each individual satellite, and to do so requires tracking the object’s course.

https://safran-navigation-timing.com/resiliency-in-pnt-gps-g...

I’m not sure if they still use it, but there was an Inertial Reference System on board - you set your location at the start of the flight, and then it can (roughly) give you a position using dead reckoning. A large disagreement between this position and the GPS position would result in an alert.

Definitely still a thing and I think that’s the goto when GPS fails. Of course there’s drift that accumulates. I think the primary threat to commercial aviation without GPS is guided landing systems where you need the location precision.

Not just near battlezones, just being near a jammy neighbour is enough: https://yle.fi/a/74-20086338

NDBs/ADF are much more common outside the US, where there isn't a huge installed VOR network or where there is a need for long-range airways over remote terrain. (Canada, I'm looking at you.)

> replacing a real radio beacon with its true position via GPS / IRS

That's the last thing you want to do. Here in Europe, we're dealing with serious issues because the Russians are jamming GPS from somewhere in Kaliningrad, but unfortunately we can't respond adequately without legitimately risking WW3.

[1] https://www.bbc.com/news/articles/cne900k4wvjo

Depends on what you mean adequately. We probably shouldn't carpet bomb them. But there's a lot more that could be done - turning the weapons donations to Ukraine from a drip-drip of the spigot to a firehose should wrap the war up.

> In general ADF is so inaccurate that its protected areas are massive

Great for listening to LW & MW(AM) radio stations in the cockpit though. Rumour has it some pilots would patch big sports games into the in-flight entertainment channels so you wouldn't miss them just because you're in the air.

Easiest to understand with a flashing + rotating light analogy.

When a rotating, say one turn in 10 seconds, green light is pointing north, a red light flashes.

If you see a green flash 1 second before a red flash, you're on the 36 degree radial.

Wouldn't you be on the 324 degree radial

Depends on the direction of rotation which the GP didn’t specify.

Here is the authoritative video of an English radio engineer explaining how VOR works.

It truly is 60s level wizardry.

https://youtu.be/tZ2gG1v9Xg8

If you want some in-depth technical details on how VORs work, I found this video very helpful. It goes into many of the details of the analog radio engineering:

https://www.youtube.com/watch?v=tZ2gG1v9Xg8

You can also compare and contrast how VOR/DME and TACAN works, to see two different solutions to one problem (polar coordinate measurement in 3D space.)

The mechanical differences are quite obvious when you look at a VOR, a TACAN, and a VORTAC, but the engineering behind them is interesting.

Animats wrote a great comment on the topic here [1] as well.

[1] https://news.ycombinator.com/item?id=40400619

Cute.

He set up his receiver near the VOR, though. So he doesn't get any useful distance info from it. He can hear the aircraft's query and the fixed station's reply, but near the DME station, the difference will be constant, just the fixed delay.

The next step is to put the receiver far from the DME station. Then, the time delay measured will indicate the aircraft to DME station distance minus the aircraft to receiver distance. I think this lets you locate the aircraft somewhere on a hyperbola, similar to the way GPS and LORAN work off time differences. If you have two receivers at different locations, you should be able to get two hyperbolas and locate the aircraft.

This is really a 3D problem, because altitude. So you get quadric surfaces and need 3 receivers. Preferably four, because there are multiple solutions. Two is enough to get a rough aircraft location for test purposes.

This has potential as a ground backup for ADS-B. ADS-B tells you where the aircraft nav system thinks it is. This is telling you where it really is, if it's using a VOR/DME at the moment.

But not who it is. That's not in the DME poll.

We already do something vaguely similar with MLAT, measuring the time delay from transponder signals at different receiver sites.

MLAT data can be used for either unofficial situational awareness in non-radar facilities (to display non-ADS-B aircraft), and in some limited cases can be fed directly into official radar displays when running in sensor fusion mode.

> if it's using a VOR/DME at the moment

This is the crucial thing. The article says

> The pilot will usually tune the radios to the stations that are part of the procedure that the aircraft is flying (although the pilot is free to tune to other stations as a cross check), so the kind of aircraft that we expect to see in the recording are those operating on the Madrid Barajas airport, not those flying high en route.

The article author has it right. Nowadays most aircraft are using GPS to navigate, and only use DME if on a specific approach procedure that requires it. In practice, this has far narrower scope than ADS-B.

Another commenter has it right - if instead of an experiment you actually want to locate aircraft without (or not) using ADS-B, you're far better off doing MLAT on Mode S, though you do need multiple spatially separated receivers for that. Aircraft are far more likely to have a Mode S transponder and have it switched on than they are to be using DME on the frequency you choose to monitor.

Even if an approach uses DME, depending on the aircraft or company operating procedures they still may not be using DME, because GPS is a valid substitute for DME in an approach and more convenient if you’re already otherwise using GPS RNAV.

In fact it’s quite common to shoot approaches that have DME specified fixes in an aircraft that doesn’t even have a DME transceiver.

LimeSDR is a beautiful board. FPGAs are the best interface with analog RF.

Reply on news.ycombinator.com