94 comments

[ 2.8 ms ] story [ 168 ms ] thread
Does anyone have a technical explanation of how gps spoofing works?
(comment deleted)
I'm not sure this is technical enough, but this gist is as follows. GPS signals are extremely weak. They are below the background noise floor (you can't see them with a spectrum analyzer) and can only be received by convolving with a pseudorandom code (PN code). Because they are so weak, an adversary can overpower them and cause receivers to "lock on" as long as rogue signal is well designed. GPS signals are fairly simple. They mainly encode precise time information from atomic clocks orbiting the earth. By receiving 4 time signals from 4 atomic clocks in known orbits, one's position can be precisely known.

This is a hard attack, but not not impossible. The Iranians are rumored to have used this technique to down a stealth drone the US was using for spy purposes.

I'm sure it's beyond me without a lot of work, but what makes the attack difficult for someone with the right background?
I think this has become much easier with the rise of SDR.

You would mainly have to know alot about the orbits of the GPS satellites to be able to convincingly simulate their signals from a given point on the globe.

> You would mainly have to know alot about the orbits of the GPS satellites to be able to convincingly simulate their signals from a given point on the globe.

Do GPS receivers validate satellite orbits? Do they even know what all the satellite orbits are supposed to be?

If it is indeed GPS spoofing it must be happening to commercial vessels right? Doesn't the US military use a different, more accurate/secure version of GPS? It seems more likely the less secure AIS protocol is somehow being manipulated.
Yes, the US military uses a different GPS code to the general public. It's more accurate and harder to spoof.

However, these collisions could be caused by spoofing the general maritime traffic so the US ships are on the 'correct' course and everyone else is offset.

> However, these collisions could be caused by spoofing the general maritime traffic so the US ships are on the 'correct' course and everyone else is offset.

These collisions could not be caused by anything but massive human failure, because while the military may exploit networked electronic conveniences to track expected civilian traffic positions, that's not supposed to be the only system they have.

And they'd be useless as warships if that was their only way of detecting larger ships than their own size short of bumping into them.

(comment deleted)
GPS can be selectively denied to an area (all but US military hardware) while the US is engaged in combat.

There are also military specific parts of GPS that civilian receivers can't access. I don't know if military receivers are ignoring civilian signals, though.

> There are also military specific parts of GPS that civilian receivers can't access. I don't know if military receivers are ignoring civilian signals, though.

For all intents it's a different system, the packets are encrypted, the right receiver hardware gives you vastly superior accuracy. There's civilian hardware which can use the packets without decrypting them for better accuracy, there's some patents on doing this amusingly.

Mariners don't rely on AIS for collision avoidance. It's all visual spotting, radar, lights, and horns. Everyone knows that AIS transmitters sometimes fail or get turned off. And the reported position may lag behind the actual position by several minutes.
He's just making that stuff up.

The US Navy has released surprisingly little info on the Fitzgerald collision. The new "report" is all about damage control after the collision, not the events leading up to it.

I am actually more concerned about the 10 missing sailors, not just concerns for their lives, but I am really concerned how they went missing after the collision? The ship's damage is not very bad. I can understand they could have fallen into water, but this doesn't make a lot of sense. If they were facing the impact, they would have run away. If they were in other three directions, then I am surprised that ten sailors got thrown off the deck, and none of them could get back on the water surface. This raises my concern for the ship's rails.

--EDIT--

I believe this incident occurred early in the morning around 5:24 a.m and I don't think it was in total darkness. Also, what about the transport ship? Shouldn't they also have the ability to detect (and also human for lookup) since this is one of the busiest lane?

I read in the other collision that water flooded the rooms very quickly, which makes it difficult to escape.
> The ship's damage is not very bad.

That's not accurate.

Not very bad even though water flooded in. But the ship was still in good condition since they were able to stop the flooding and isn't beyond repair.
By design, a modern warship can take an immense amount of damage and remain buoyant. That comes as little consolation to the poor buggers trapped in a flooded compartment.
The ship was hulled below the waterline; this damage is incredibly serious. The best that can be said about it is that the ship was able to return to port under its own power.

> the ship was still in good condition

This is just a crazy thing to say.

Sorry, I don't agree it is crazy thing to say. This ship is not beyond repair and is actually not hard for repair. A sinking ship or a ship on fire or a ship with half of the deck destroyed or a ship with a hole in the middle would be considered in serious condition. This ship would still be in active fighting if this ship was in the middle of a naval battle. No one would be crazy enough to abandon this ship just because some compartments were flooded and under control.
(comment deleted)
Opening flooded compartments is not done lightly, you need to stop the source of the leaks first before you can open the hatches and enter the compartments and do a full search.

A collision between a 10,000 ton destroyer and a 20,000 (?) ton cargo ship dissipates a lot of energy into the water and churns up a lot of turbulent water, if someone fell into the water in the dark, it's not at all surprising that they'd be unable to surface. A human loses buoyancy around 30 feet deep in the water so if you're pulled 30 feet down in the dark, it can be disorienting and hard to figure out which way is up.

Is it a speculation suggesting some if not most are actually in the flooded compartments? Perhaps I didn't see this part in the articles I have read so far.
Since the missing crew are still missing, it's speculation that any are in the flooded compartments, but some of the flooded compartments are crew berths:

Monday's collision east of Singapore was the second involving a ship from the U.S. Navy's 7th Fleet in the Pacific in two months. The Navy said "significant damage" to the McCain's hull caused flooding in adjacent compartments, including crew berths, machinery and communications rooms.

It is speculation, but unfortunately it's also the most likely scenario.

    - Crew berths were at the primary impact point, so are almost certainly sealed at the bulkheads by damage control.
    - Incident occurred at 5am local, when you'd expect berths to be popular.
    - Sunrise is 7am local, so it was dark - so you wouldn't expect to have any unnecessary crew topsides (precisely because finding MOBs in the dark is near impossible) - and you'd hope any that are, have protective or flotation equipment.
    - You'd also expect that 10 people running around on deck would have had a much better chance of spotting another ship.
    - The proof is in the pudding - for not one of them to have been spotted in one of the world's busiest corridors, doesn't bode well.
> A human loses buoyancy around 30 feet deep in the water

Learned something new today, thanks for that. Didn't even realize this was possible.

Learned something terrifying today, is more like it. No wonder those big wave surfers talk about being unable to tell which-way-is-up when they get washed up by a wave.
The science behind it is really interesting. I had always assumed that you'd get more buoyant as you went down and the pressure got higher; the thing is, buoyancy isn't about pressure as such, it's about relative density. Water is almost completely incompressible; even miles down, its density only increases by a few percentage points. But the main thing keeping humans buoyant is the air in your lungs, and your lungs are compressed as the outside pressure rises, making the airspace smaller and denser, until your average density is greater than the water's and you sink.
Yeah, I looked it up for the same reason, that's why I said I found it so interesting :) thanks for mentioning it though, I'm sure people will be interested to learn why.
Thank you for the explanation. This is entirely counter-intuitive to me.
Regardless of any GPS spoofing that could be present, it'd be awfully hard to hide the radar return from a large tanker/cargo ship.
All ships should be keeping a physical lookout at all times, it can only be complete incompetence at work. Pretty pathetic of a warship cannot steer clear of such a large vessel. Clearly it would pretty easy for an enemy to attack them if they are thus incompetent
> All ships should be keeping a physical lookout at all times

I know nothing about the military or naval practices but this sounds hard and very costly... you want someone on shift 24/7 with his head constantly rotating watching out in all directions? Or maybe 3-4 people simultaneously looking out each in a single direction? That would mean you have a bunch of people doing literally nothing productive (besides watching) at any given point in time, some of them even being asleep since you want 24/7 shifts... and they have to not get bored or look away for any nontrivial period of time. It seems very difficult.

a bunch of people doing literally nothing productive (besides watching)

They are not doing "nothing productive", they are keeping watch to prevent exactly this type of accident from happening.

Do you really think that the entire ship goes to to sleep at night on a 2 billion dollar naval destroyer while it cruises in a crowded shipping lane? Even the $250M office/residential building where I work has 2 night watchmen and a building engineer or two on duty at all times... and the building is unlikely to have a collision when another building moves into its path.

> Do you really think that the entire ship goes to to sleep at night on a 2 billion dollar naval destroyer while it cruises in a crowded shipping lane?

I didn't say they all go to sleep, they probably have lots of things to do at various times. And I wasn't talking about more-sensitive points like crowded shipping lanes. I was just talking about having people on duty to literally keep a watch out over the sea the entire time. Not sure why you twist my words like that.

Or you're saying they do have watchmen 24/7 but their watchmen are just blind? How did they manage to collide 4 times this year if they had people watching the sea the whole time?

Oh, and don't forget others are suggesting sleep deprivation as the cause. Which is not exactly inconsistent with my point that keeping watch is probably difficult and costly...

The International Regulations for Preventing Collisions at Sea (1972) says in part:

Every vessel must at all times keep a proper look-out

by sight (day shape or lights by eyes or visual aids),

hearing (sound signal or Marine VHF radio)

and all available means (e.g. Radar, ARPA, AIS, GMDSS...)

in order to judge if risk of collision exists

https://en.m.wikipedia.org/wiki/Lookout

That is exactly what international maritime law requires. For reference, I estimate the eye height of people on the bridge and watch deck is about 28.5m above the waterline (from a picture of an Arleigh Burke-class destroyer on wikipedia), at that distance the horizon is approximately 19.1km away or a ship higher than that come into view from further away depending on the height of that other vessel which would be substantial for an oil tanker of cargo vessel (which has masts with radar reflectors and navigation lights). Now an Arleigh Burke-class destroyer is listed as having a greater than 30knot (56km/h) top speed on wikipedia, this probably means it is less than 40knot (74km/h), however it is probably not cruising at more than 56km/h so lets assume that is the common cruise speed (which is fast for a ship). So at 56km/h 19.1km takes 20.6 minutes to cover even if the other vessel is going the same speed head on that still gives more than 10 minutes to avoid a collision from when the other vessel was first able to be spotted.

So WTF was that destroyer doing for 10-20 minutes minimum that had it in a collision with another ship?

I am not a naval person by any stretch, but I thought the protocol on any large ship (military or civilian) was to have at least one, if not two people on watch at all times?

I would assume a navy ship would have multiple people with eyes 'outside the boat' especially in a busy shipping (and piracy ridden) lane such as the Malacca Straits?

US Navy ships underway typically have at least 4 crewmen on watch as dedicated lookouts, plus at least 4 more on the bridge sometimes looking out the windows and someone in CIC watching radar. In high risk situations there may be even more.
This is purely uninformed speculation. Radar and eyeballs are sufficient to prevent these types of incidents.

Sleep deprivation has continued to be pointed to as a major issue on ships, and yet from the CNO down, this has been ignored and nothing done: https://www.usni.org/magazines/proceedings/2017-07/let-our-s...

I wouldn't be surprised if sleep deprivation was a leading cause. The military in general has a problem with letting people get enough sleep, especially if the personnel are not on the day shift/watch.

With two deadly collisions in the same fleet, I bet there is some silly regulation/task that is keeping watch standers from getting enough sleep. It could be as simple as letting sleeping personnel get pulled into working parties as everyone else is on watch or doing repairs/maintenance.

I wouldn't be surprised if things got worse as they pile even more training requirements on trying to fix the issue.

Even if they were spoofed, every Navy ship has many people on watch at one time and radar is more than enough to compensate. A more reasonable explanation is exhausted crew and human error.
... which in turn would imply unreasonable schedules and systems which are not robust when faced with inevitable human error. In other words, failures of high level management and system architecture.

I await the scapegoating of the crew to protect the higher-ups.

> I await the scapegoating of the crew to protect the higher-ups.

Considering the leadership of the Fitzgerald have been removed, why would you expect scapegoating of the crew to occur in this case?

> The commanding officer, executive officer and senior non-commissioned officer of the USS Fitzgerald have been removed from their duties for cause amid the fallout surrounding the deadly collision between the USS Fitzgerald and a cargo ship off the coast of Japan on June 17.

"We've lost trust and confidence in their ability to lead in those positions and they will not return to the ship," Vice Chief of Naval Operations Adm. William Moran told reporters at the Pentagon late Thursday. The 7th Fleet also said several junior officers were relieved of duty.

http://www.cnn.com/2017/08/17/politics/uss-fitzgerald-leader...

Do ship commanders such as the removed captain of the Fitzgerald decide staffing levels? If sleep deprivation contributed this accident (as many are speculating), I am nevertheless skeptical that the people who are responsible for systemic lack of sleep will be held accountable.
You mean the speculation mentioned in comments here? This article, which itself is unsubstantiated speculation, makes no such mention. If you're just spitballing, fine, but I don't see any reason to make statements like these without doing some research first, much less express skepticism, particularly when presented with very recent evidence that leadership is being held responsible.
>Do ship commanders such as the removed captain of the Fitzgerald decide staffing levels?

And do they determine "deadlines" like when they must arrive at a destination? How much flexibility do they have in adjusting arrival time, particularly if [unforeseen] circumstances delay transit?

Twofer: Not only deflect blame from actual problem but also raise funding for anti-cyberwarfare boondoggle!
Suddenly a 500k PR contract from the Navy lands in his lap.
The dudes at CrowdStrike are all, "why didn't we think of this?"
> The U.S. military uses encrypted signals for geolocation of vessels, rather than commercial GPS.

It seems like a massive oversight to keep encrypted GPS signals for use exclusively for US military. As mentioned in another comment, the collision could have been caused by spoofing the commercial ships GPS rather than a cyber attack on the Navy vessel.

The ability to selectively enable GPS in a region for only US military (as mentioned, once again, in another comment) doesn't even seem like that big of a strategic advantage - surely a sophisticated enemy wouldn't rely on GPS, would they?

Any strategic advantage doesn't really seem worth the potentially massive damage that could be caused by a large scale spoofing of GPS signals.

Plus with the advent of Glonass, Galileo, and Beidou, the ability for the USA to turn off GPS has less strategic value.

https://en.wikipedia.org/wiki/GLONASS https://en.wikipedia.org/wiki/Galileo_(satellite_navigation) https://en.wikipedia.org/wiki/BeiDou_Navigation_Satellite_Sy...

That would be true if the US did not have it was not willing to use effective ASAT weapons to also deny opponents the use of GPS alternatives in the event of a war in which they would want to deny GPS.
Surely GPS satellites can also be shot down, no? It seems the best possible outcome might be a single global navigation system that all powers depend on, creating a navigational form of mutually assured destruction, ensuring its continued availability during a time of conflict.

That will probably never happen, though.

I feel like that would just encourage all parties involved to develop robust non-satellite fallback systems and procedures...which, actually, would have prevented these collisions in the first place.
All parties involved already have robust non-satellite systems for collision avoidance because satellite systems have never been used for that purpose in the first place.
The post suggests that they weren't robust enough. That's why I said systems and procedures--backup systems don't work if the guys monitoring them aren't well-trained or paying attention because everyone assumes the GPS is infallible. Policies based on GPS MAD would place a lot more attention on the non-GPS systems.
Seems more likely that they'd use local jammers where needed than to risk global retaliation against the USA by shooting down 10's of billions of dollars of European, Russian, and Chinese satellites.

Assuming, of course that they aren't fighting a common enemy which would let them convince all of those satellite owners to shut them down or degrade the voluntarily.

> surely a sophisticated enemy wouldn't rely on GPS, would they?

That leaves the unsophisticated enemies.

That's nonsense, there's no way that some unknown third party could have tricked the Alnic MC into an allision with the USS John S. McCain. It's the sort of thing that only works in movies, not real life. Mariners don't use GPS for collision avoidance. This is a simple case of incompetence by the crews of both ships.
I imagine one could make a modern device that tracks position by watching the sky. I wonder how well could that work in bad weather. Is there any light spectrum of the stars that passes through a serious storm? Humans may be incapable of seeing it but how about advanced telescope/camera?
A much better explanation for that is that the guys there are being overworked and exhausted. Navy resources are overstretched as there is lack of destroyers due to many years of interruption in DDG construction. Navy has at least 12 destroyers less than it would have if not the catastrophe of Zumwalt class construction.
Navy resources are mostly stretched because of the Navy's insistence on launching $30 billion supercarriers to fly $120 million F-35s. You could build 15 or 16 Arleigh Burke-class destroyers for what it's costing (so far) to build the Gerald Ford.
That is just because it is the first ship in class. Neither Enterprise nor Nimitz came out easy either.
This article is just idle speculation with some of the vague 'Russia is behind it' rhetoric we've been seeing over the past few years.

Even if someone had developed the capability to spoof military grade GPS why would they use it on a random ship for no reason? That would accomplish nothing while tipping off the military that someone had cracked GPS.

Even if this is a GPS problem at all it is more likely to be a bug in the implementation than a targeted attack.

By Occam's Razor this is likely just a plain old human error and not some sophisticated conspiracy - someone was negligent and shit hit the fan.

Could they instead spoof the GPS that the other civilian ship was using and guide it into the US ship? That seems like it would be infinitely easier to do.

Isn't is also possible to just hack into the computers of the civilian ship as they are likely incredibly insecure and basically do what you want with it? I understand most industry computer control systems incredibly dated and insecure.

Just speculation of course.

Navy attack ships are much, much faster and more maneuverable than cargo ships or takers. They could have outmaneuvered one easily if they wanted to.
The US military uses encrypted GPS, not the civilian one.
Maybe, but why do it? I see no possible motive here.
I agree, that makes little sense of giving away such a capacity outside of an armed conflict. Anyway, there were 4 collisions in this area this year (we mostly talked about the 2 with loss of life) so there may still be a more fundamental problem to solve than individual negligence.

Just to play devil's advocate, here are a few hypothesis that could make malice have more sense in that case:

- The enemy has the capacity to spoof this but knows that it will soon be removed. May as well use it.

- The enemy knows that this vulnerability will not work on ships that are in conflict mode and is interested in removing two ships from the existing fleet in preparation of a conflict.

Pretty weak hypothesis if you ask me. Anyway, I can't say I am sad at seeing media and the general public talk a bit more seriously about cyber security.

Especially since USN are making these mistakes with increasing regularity: Port Royal, Porter, Antietam, Fitzgerald, McCain, etc. International shipping predates GPS by centuries. It's entirely possible to navigate and avoid collision without using GPS at all. The Navy (or more probably, a particular subset thereof) are just increasingly bad at driving their boats.
Unfortunately I see 'complete incompetence' as the only rational explanation. This is just the latest in a string of naval disasters this year which are not just confounding and bizarre rank amateur displays, but have resulted in loss of life.

At this point, the American people deserve nothing less than an extraordinary response and complete accountability throughout the ranks. When destroyers are getting rammed by cargo ships, and guided missile cruisers are colliding with fishing bots, and running aground, we have a systemic problem with training, discipline, or protocol, likely all three.

You simply cannot blame broken GPS on getting your destroyer rammed by a 20,000 ton cargo ship. Our enlisted deserve better, and Admirals should be getting shit-canned for this.

I vaguely recall a story, maybe a decade or two back, about a naval wargaming exercise in which the red team (simulating the enemies of the US) approached a zillion-dollar warship in rubber rafts that didn't trip any of its cutting-edge electronics and "disabled" it with nothing but some men and some simulated explosives.

They were accused of cheating, the officers upbraided, and the whole scenario rewound to erase the embarrassing incident.

Out of curiosity, how doe they go up the ship when they approach it with a raft? Do they manage to tie a rope around something in the deck somehow, or do they cut a hole in the ship...?
I don't remember the details, unfortunately. They may have just planted (simulated) shaped charges directly on the hull, or they may have gone through some rigamarole with grappling hooks or something.

It occurs to me that you could probably climb the outside of a ship with sci-fi electromagnetic gloves/boots, but I've never heard of anyone actually making anything like that.

Normally warship hulls are degaussed to remove magnetic signature for mines. that might stymie any attempt to climb using magnets.
It's hard to find good information on the simulation, but the reference appears to be to the following: https://en.wikipedia.org/wiki/Millennium_Challenge_2002

The NYT reported many years later: "In the simulation, General Van Riper sent wave after wave of relatively inexpensive speedboats to charge at the costlier, more advanced fleet approaching the Persian Gulf. His force of small boats attacked with machine guns and rockets, reinforced with missiles launched from land and air. Some of the small boats were loaded with explosives to detonate alongside American warships in suicide attacks. That core tactic of swarming ..."

Thus, to answer your question, they did not go up the ship. The attack is generally reminiscent of the USS Cole bombing, which preceded the simulation by a couple years.

The crew uses cellphones (browsing Singapore Tinder is a really good way to kill time while not falling asleep during the watch at 4am - makes for a peculiar case of "texting while driving" though). Coming from the providers in the region, with many directly or indirectly owned/invested/hacked by China, the phones are already hacked, and they provide a pretty good attack vector on any wireless ship system and after that on any other that it is connected too. Though my bet here is on pretty obvious Chinese spoofing of the destroyers' GPS - it being supposedly military grade makes it the last thing to be looked at.
Do you have any evidence that any networks accessible by personal crew devices (i.e. cellphones) are connected at all to the ship's control networks?
are you kidding? Check out "smart ship" for example:

http://www.globalsecurity.org/military/systems/ship/ddg-51-u...

Everything integrated, and as long as at least one component talks wireless ... and there are actually a bunch of them, and as i mentioned, being an actual smartphone is only one of the features of those cheap smartphones bought at the local port :).

One can easily imagine that the ships which rammed (and/or some other ships close by) the destroyers could have carried cell towers and/or GPS spoof-er which were promptly, say, dumped overboard.

And leaving all that hi-tech aside - plain old $2K/bottle cognac, drink of choice of the 7th fleet :) https://federalnewsradio.com/tom-temin-commentary/2017/03/fa... , http://www.washingtonpost.com/sf/investigative/2016/05/27/th...

I don't see anything there that suggests that the ship runs on a wifi network.

The closest thing I saw was "Shipwide interior wireless communications system", but that doesn't imply Wifi, and is likely just a digital trunking radio system -- not something a consumer Wifi device will be able to connect to.

"Cyber" seems to have taken on the role formerly filled by angry divine beings: If something happens and we don't know what it is or why it happened, we used to think someone angered a god or an ancestor. Now, it's cyber.

Don't forget that several wars have started based on misattributed events. In just the U.S.: The Maine and the Spanish-American War, the Gulf of Tonkin and the Vietnam War, Iraq's WMD program and al-Qaeda alliance and the Iraq War ...

And attribution is especially difficult in attacks on computer systems. An enterprising, well-resourced actor could provoke war between enemies or division between allies with a well-crafted attack.

Don't forget six years ago when we destroyed Libya because some exile in Switzerland claimed Qaddafi was being extra super mean, only he wasn't. How long until we admit that Assad wasn't connected to the Syrian gas "attacks"?

"Misattributed" is soft-pedaling it. The war pigs tell these lies so they can make money by killing hundreds of thousands of innocent people.

> we destroyed Libya

The U.S. and allies helped drive Qaddafi from power. They didn't physically destroy Libya, though maybe politically.

> because some exile in Switzerland claimed Qaddafi was being extra super mean, only he wasn't.

Qaddafi's human rights record was very well known for a long time, and he also made enemies in the West by supporting terrorists and destroying a commercial airliner in flight, killing its passengers. The immediate cause of the West's intervention was that Qaddafi was openly about to destroy rebels (including civilians, IIRC). I'm not sure how much of a role this exile played.

Educate yourself:

https://www.foreignaffairs.com/articles/libya/obamas-libya-d...

http://foreignpolicy.com/2016/03/22/libya-and-the-myth-of-hu...

Holy shit those lies were so transparent that even staid mainstream media brands felt compelled to debunk them! Even if those lies had been true, why did we care so much about Libya and not about nineteen other brutal corrupt dictators in Africa and around the world? Qaddafi had already given up his nuclear program, and DPRK rightly points out that that mistake cost him everything. How convenient for those who benefit from both the current European refugee crisis and from escalated conflict in Korea!

The U.S. government's budget sequester has forced the military to greatly reduce training and therefore readiness. The Air Force attributed recent fighter plane accidents to a lack of training. I wonder if these accidents are related.