Air travel is not ready for electronic warfare | Hacker News Reader

Air travel is not ready for electronic warfare

jeffwise.net

43 pointsano-ther50 comments

The author seems to believe that GPS is the only and primary form of navigation.

Pilots in the US are trained from day one to use landmarks and dead-reckoning. From there the skills are built upon to where I believe it would be difficult to push a crew to such a gross navigational error that it causes disaster.

I’ll need to read the article more closely but it appears to be a lot of fluff and leaves out the many layers of navigation still available.

It's mostly fear mongering for Regular Joe, along with:

> Another important development was Russia’s invasion of Ukraine in February 2022. Vastly outnumbered in jets and tanks, Ukrainians quickly jury-rigged a miniature air force out of off-the-shelf drones carrying improvised munitions

Lol, Ukrainians were watching videos from Syria and Afghanistan and implementing ISIS tactics for years before 2022.

I think the drone war is getting off the charts. 300,000 per month!?! [1] This will definitely have impacts in the coming decade both in war and terror/counter-terror applications.

[1] https://bulgarianmilitary.com/2024/01/05/russia-produces-300...

Commercial traffic flying at 30-40 thousand feet is not flying VFR typically. And of course spotting landmarks through cloud cover or at night is not a thing. But you are right that spoofing would largely just be annoying. Commercial traffic is usually in contact with a controller and monitored on radar. So, any course deviations would result in controllers asking what they are doing and be followed by them simply vectoring them back on course with compass headings.

However, GPS has replaced a lot of the NDB and VOR radios that pilots used to rely on. Also a lot of airport approaches are now GPS rather than ILS/VOR. It's pretty critical to modern aviation.

GPS, or rather GNSS (which includes GPS and other systems), spoofing is a concern; but so far mainly in conflict areas. Most airlines would route around those.

Resilience against spoofing is actually a big topic from a strategic point of view. Also, spoofing is a big part of counter measures against drones in e.g. Ukraine. So there is a bit of an arms race going there.

The point with GNSS is that most phones can utilize multiple satellite positioning systems at the same time at this point. Which is of course something that navigation systems in planes could also do. And of course many pilots actually use ipads in the cockpit, which they could use as a backup in case of an electrical failure in the plane.

I’m well aware of the methods used for navigation and the technological changes over the last decade or two.

My point is that US trained pilots have an exceptionally strong foundation of the fundamentals and are very unlikely to fall prey to a misguided GPS signal. The spidey senses will be tingling is essentially my argument.

A generation or two from now that may become more of an issue as primary training is beginning to move away from the fundamentals and is bracing technology.

For the past 10 years I've been travelling to Russia occasionally. Normally during a flight I like opening up my GPS app to check the location, altitude and speed of the aicraft.

About a year ago I noticed that shortly after crossing Russias border the GPS stopped working. Initially I thought it was a glitch, but after a few flights it became obvious that something was going on, probably the signal was baing jammed.

Interesting that close to landing when the altitude is lower than 200m, the signal is restored.

It's unclear to me how the plane is navigating though if the GPS is jammed. Very weird...

GPS isn’t the only navigation tool. INS is pretty accurate if you zero it correctly, and VOR is good enough to keep to a specific airway.

All INSes require periodic GPS updates to limit the growth of positional error. Usually it's once per second or once every 10th of a second.

That sounds way too frequent, isn't typical holdover like 12 hours?

This? https://www.defenseone.com/defense-systems/2023/12/russia-ja...

As for planes, can they do GLONASS?

I would imagine any military “GPS Jammer” would also block similar systems…

My phone is supposed to support GPS, Glonass, Beidou and Galileo, so apparently all of them were jammed.

I guess it would make sense if they were all jammed, though I'm not sure the fact that your phone didn't pick those up really proves this was the case? I have no idea how these work but I could totally imagine phones interpreting jammed signals differently from weak/nonexistent ones. Plus, I've had at least one that nominally supported all of those but in practice didn't really feel that way. You're probably right, just not sure you can deduce it from your phone in particular.

Planes have used inertial navigation for a long time. Adding GPS to that mix is relatively recent. GPS was never really intended to be used for navigation even though it can be and often is. There was no need for it to be particularly reliable when it was designed since the Soviets could easily kill it if they wanted to. It had a different purpose.

Can you elaborate on what you mean by "not intended for navigation"? GPS was intended as a replacement for TRANSIT and LORAN, which have navigation literally in their names. It's also the textbook example on Wikipedia's satellite navigation page, and I can't imagine how precise geolocation somehow wouldn't be useful for navigation.

Also, the general analysis after KAL007 was that GPS should be made publicly available specifically to aid civilian navigation. They even identify what we call RAIM today to make it suitable for eventual usage as a primary navigation system.

GPS was trivially attackable from inception, that was not a design requirement given its purpose. The purpose of GPS was to precisely measure the planet in peacetime. A unique core capability of the US military is extremely precise INS technology, the capabilities and precision of which are closely guarded secrets and under continuous R&D. The precision of INS is dependent on the precision of your world model. GPS allowed extremely precise world models to be built globally, which provided the data model INS needs to be maximally effective. In wartime the GPS system can be killed by any near-peer adversary, but the damage has already been done to the extent inertial targeting systems have a precise model of the world.

No US military system relies on GPS for navigation. Military navigation systems use GPS as an untrusted source for fine-tuning inertial navigation within tight error bounds. Against US military systems, successfully spoofing GPS might buy you several meters of deviation. This only affects the cheapest US guidance systems, since most weapons have active terminal guidance. It is worth noting that GPS corrections are being phased out in US weapon systems, purportedly due to improvements in INS tech that moot the value of GPS corrections.

The tl;dr: GPS was developed to build a precise model of the world in peacetime that could be fed to inertial targeting and navigation systems in the complete absence of GPS. In that, it has been a massive success. US military systems have never relied on GPS for anything important. Contrary to popular media, the US has never produced GPS-guided weapons, even the cheapest systems are INS at their core.

Civilian systems use GPS for navigation, against its design, because they can mostly ignore its trivial susceptibility to hostile actors.

Most of the basic attacks against GPS also worked against TRANSIT, LORAN, OMEGA, and others, which were clearly used for military navigation. I'm not seeing the design differentiation with the systems that were guiding ships and missiles.

All US military navigation systems use INS pervasively for primary navigation. You can attack GPS, LORAN, etc and it won't affect their operations. This has been axiomatic in the design of those systems for many decades. US military ships are no different, also using INS for primary navigation. As with every other US military system, they can conditionally accept fine-tuning inputs from GPS and other untrusted sources of navigation data because why not. They use a lot of open source data sources, but they don't need them or trust them.

> It's unclear to me how the plane is navigating though if the GPS is jammed. Very weird...

RNAV (or aRea Navigation) based on VOR-DME has been a thing since (at least) the 1970's. You put in a waypoint based on a radial and distance from a navaid, and then the box in the cockpit would Do Math so you could follow a track to/from that virtual navaid.

Bendix even made a cute little box for small airplanes in the 1980's that allowed you to pre-store up to 4 (or 10) such waypoint-frequency tuples and cycle through them to approximate a straight path from your origin to destination, using navaids near (but not on) your route. Given the service volume of a typical VORs, the 10 waypoints could enable a flight segment of over 2,000 km before you needed to enter more waypoints.

Around the same time, the flight management systems in airliners began to have navaid databases, allowing you derive lat/lng from the angle and distance. They could even auto-select the best local navaid to use. That enabled DME-DME area navigation, which has ~10x better precision than VOR-DME.

The flight management systems in modern airlines of course still have these databases and support navigation by ground-based navaid.

Russia still operates their version of LORAN (Chayka), too, so that's another option. By the 1980's, LORAN receivers were pretty well automated and directly produced lat/lng outputs. LORAN has worse accuracy than GPS, about ~400 meters, but is still good enough to get you to the airport.

And, as several other people have mentioned, virtually all airliners continue to carry an INS.

If you know GPS is unreliable in specific areas, it's not hard to get around with rough accuracy. There's still a compass, after all.

I recently flew Lufthansa over that area. Here's how the inflight map looked like https://i.ibb.co/ryFdXyC/IMG-8587.jpg

Fortunately I only saw this afterwards, otherwise I would have been worried.

(obviously we didn't actually make this detour)

>> President Reagan ordered an unencrypted version of GPS to be made available to the public so that civilian aircraft would have a reliable form of navigation. At the time, no one apparently foresaw any motive or incentive for tampering with GPS signals, so no provision was made to protect them.

That can't be true, I believe one of the Die Hard films made at the time is based on this idea.

Selective Availability was baked in from the beginning.

It would probably be a good idea for Starlink in-flight WiFi to provide position information, so pilots could use a phone/laptop as an emergency backup.

Starlink is harder to jam than GPS, because the signals are strong and directional.

There is precedent for using consumer tech in an emergency. After the Qantas 32 explosion, the pilots used cell phones to contact the airline because most of the plane's radios were dead: https://www.youtube.com/watch?v=DS9rYWfdC9w&t=1577s

Perhaps I just don't understand how this works all that well, but shouldn't it be easy to defeat GPS spoofing when you're 30,000+ feet in the air? Presumably the spoofed GPS signals are all coming from the ground, or close to it. I would think it would work to use a semi-directional antenna that is on top of the aircraft, pointing upward. Even an antenna that has a 180-degree field of view should reject fake GPS signals from the ground, no?

Only thing I can think of that would invalidate this is if the fake GPS signals bounce off the atmosphere above the planes and come back down. What else am I missing here?

the gps receivers are probably not going to be able to detect the direction a signal comes from.

You're right, a cheap GPS isn't going to do that.. however, it is well within then realm of possibility to design a GPS receiver system that uses beamforming to lock onto satellites and ignore everything else for orders of magnitude better jamming resistance.

Its called controlled reception pattern antenna (CRPAs), which unfortunately is an ITAR restricted thing because of insane paperwork required, so no one wants to build it for civilian receivers. Which is stupid because beamforming (or null-steering) algorithms are not that difficult to implement, and so we already have Chinese and Iranian receivers with CRPAs.

So PNT National Advisory Board is trying to convince US government to stop putting CRPAs under ITAR: https://www.gps.gov/governance/advisory/meetings/2023-05/mur...

I figured there was going to be some ITAR thing if I kept going with the brainstorm.... and you confirmed my hunch.

There are so many things these days which are almost trivial to build and somehow fall under ITAR.

Radio goes through and reflects off things, including aircraft. GPS signals are incredibly weak, and even if an aircraft attenuates or reflects 99.9% of the signal, a ground based system will nearly always be able to overpower it.

It can be as simple as reflecting off the tail section of the aircraft.

Incredibly weak is correct and even bordering on understatement. They are below the noise floor by about 50dB (which is a logarithmic scale, so is a factor of 100,000x).

It still blows my mind that gps even works given how far below the noise floor it is. I am very far from being any sort of expert in radio theory but my layman understanding is that the receiver "knows" exactly what a gps signal should look like and somehow magically uses that to extract the data out of the noise. It is one of those absolutely genius aspects of modern life, and we tend to just take for granted

Same. I am pretty comfortable as a programmer and that translates to feeling moderately comfortable at gluing together blocks of digital electronics. Analog electronics are much more of a struggle and RF is entirely black magic. The idea that GPS works at all defies my understanding of what ought to be possible.

Maybe the GPS hardware (firmware) or software reached the conclusion it shouldn't finish calculating the result for some regulatory reason?

> Presumably the spoofed GPS signals are all coming from the ground, or close to it.

I think that's a good presumption when operating in, say, Iraq or Afghanistan. Maybe less so when operating near Russia (which has its own GNSS system)?

> I would think it would work to use a semi-directional antenna that is on top of the aircraft, pointing upward. Even an antenna that has a 180-degree field of view should reject fake GPS signals from the ground, no?

GPS receivers already do this to reject ground reflections that could give false position readings (and the GPS signal is circularly-polarized so it can reject odd-ordered reflections off building and the ground). However, any kind of steerable (electronically or otherwise) directional GPS antenna is considered an ITAR no-no.

The other problem is that GPS signals are incredibly weak (actually below the thermal noise floor), and rely on processing gain from the CDMA chipping to raise the SNR to something useful. So even an antenna with very very good rejection from the lower hemisphere will easily pass a jamming signal from the ground which is both many times closer and probably of higher power.

> Only thing I can think of that would invalidate this is if the fake GPS signals bounce off the atmosphere above the planes and come back down. What else am I missing here?

Any country with a space program could transmit the jamming signal from above.

A perfect use case for public key cryptography. Digitally sign the GPS coords!

Jamming is akin to a Denial of Service attack. It disrupts the channel. Cryptography cannot solve it.

The article and GP’s proposal both address spoofing, not jamming.

https://en.wikipedia.org/wiki/Selective_availability_anti-sp...

For those interested in a good writeup on the subject, this explains alot of why Air Travel is vulnerable:

https://medium.com/@angelinatsuboi/detecting-hacker-aircraft...

Also, here's a link to the github:

https://github.com/ANG13T/fly-catcher

A big reason for not updating avionics to more secure versions is price. Anything that is qualified to fly is very expensive.

The evidence shows that the contrary. Has a plane ever gone down from “electronic warfare?” If a motivated adversary know you have a defence, they will find a way around it, meaning you can only go so far in defence

>The evidence shows that the contrary. Has a plane ever gone down from “electronic warfare?”

Does the fact that no one has killed me yet indicate that I'm immortal?

No, but assuming most of your friends and family haven't been murdered, we may assume murder is unlikely to be the cause of your death.

Whether or not my friends and family have been murdered shouldn't be correlated with whether or not I've been murdered. Unless we're all in the same place most of the time, I guess?

I'd say it's an indicator.

For instance: You may hang out with a crowd who is inclined to engage in dangerous criminal activities. You may live in a place with high crime. You may live in a place where most people are left alone, but there are a few families with Montague/Capulet style feuds that the police have chosen to leave alone while they kill each other lol

So if I have no other information, I would say someone who has a lot of family and friends who get murdered is more likely to be murdered than someone who doesn't.

But, sticking with this analogy, several of my cousins have recently been shot while going about their day.

> Has a plane ever gone down from “electronic warfare?” I

What do you mean?

Obviously Iran used electronic warfare to capture a drone - https://en.wikipedia.org/wiki/Iran%E2%80%93U.S._RQ-170_incid...

Has a non targeted civilian plane crashed from the result of electronic warfare in the area? Not that I known of.

Close calls - https://asrs.arc.nasa.gov/docs/cb/cb_473.pdf but I guess as you claim checks and balances kept them alive.

The title is "Not Ready" so it's a call to action as this stuff ramps up.

It's not just crashes. If you have a scare you are not prepared for then airlines avoid air space that's a big deal.

The other way to look at it: there are old planes still running that seem unlikely to have an installed GPS. Air flight pre-dates GPS.

This comment seems to imply we should never prepare for future risks because they haven't happened yet. Bizarre.

Reply on news.ycombinator.com