Comment by cromwellian | Hacker News Reader

Comment by cromwellian | Hacker News Reader

Do we really need manned fighter planes at all? For a trillion dollars, you can overwhelm enemy air defenses with cheap, primitive drones in a saturation attack.

A tomahawk missile costs about $575k per unit, that lets you buy about 2 million tomahawks for $1 trillion. If you built a drone version that could loiter over an enemy airspace and then attack, sort of a tomahawk version of the Predator, you could buy a hundred thousand of them even if they were 10 million a piece.

You could also built UCAV fighters that could out-turn and fly most fighters by avoiding the need for the cockpit overhead and restrictions in g-force. Missiles make 40g turns, and drone airfames have been build to sustain 12-15g turning.

Are we spending a trillion dollars because the AirForce has a romantic notion of a human dogfighter in the seat when a guy with a joystick could do the job? Top Gun is a lot less interesting if robots or teleoperation is in play, but do we really that kind of engagement?

> For a trillion dollars, you can overwhelm enemy air defenses with cheap, primitive drones in a saturation attack.

For a few hundred dollars I can jam your communications channel(s) to the drones. Asymmetric warfare sucks.

And, depending on how your jammer works, I can probably kill it for a few thousand dollars (+- an order of magnitude).

Pentagon procurement is a shitshow, but the DoD has many brilliant radio, radar, and signals guys working for them directly, or indirectly.

> For a few hundred dollars I can jam your communications channel(s) to the drones.

The same could be said for manned fighters. Granted, you still have a pilot who's there and capable of making decisions independently of communications.

But, I imagine if you threw up enough interference to screw with a drone, you'd probably make some very pretty lights happen on an F-35's radar.

> For a few hundred dollars I can jam your communications channel(s) to the drones. Asymmetric warfare sucks.

The problem is that once we're talking tens of millions of dollars and up -- no, you can't.

We're talking the swarm is likely operating in a shared point-to-point link mode, either using highly directional antennas that are essentially deaf to below, optical links, or (more likely) a mixture of both modes.

At 10,000 drones, they likely can form a stable mesh network over the city covering a wide enough area at a few miles in the sky that you're talking about having to run power plants or detonate magnets with high explosives to generate the kind of wattage needed to shout in to the directional communications gear.

Jamming a few hobbyist drones is nothing like jamming thousands of military grade drones, and the techniques you think will stop their communications aren't all that effective.

However, scanning the area under your bot swarm of amateur attempts to jam the network is probably a great way to target RF engineers in preparation for a ground invasion, and minimize their ability to improvise communications gear once the main communications networks are knocked out.

tl;dr: Taking out a $10,000,000+ swarm of military grade weapons isn't the same as disabling a few hobbyist trinkets, and may get you seriously hurt in the process.

> The problem is that once we're talking tens of millions of dollars and up -- no, you can't.

Sure. If you're willing to build a new GPS system, because that's decently jammable for cheap.

Also are you proposing that all the drones will be individually piloted on missions as they are now? If so, I don't strictly have to jam the drones on the battlefield, I can just jam your satellite uplink or physically attack your operations centers. Add the cost of a plane ticket.

So far, we haven't used these weapons against any military that could be considered our peers. I wouldn't be so confident in our operational deployments without anything to base it on.

You've done an exceptional job of proving you have no idea what you're talking about. (Hint: I write software for a govt. contractor that builds drones and their control stations)

Can you point us to any public information about military systems' jamming resilience? E.g. here's a simple comm system, here's how you jam it, and here's a better comm system that can't be jammed that way?

I can't really get into the details, I apologize. I can tell you that the specific topic of jamming isn't a concern. Try and think about UAVs in the same way you might think about commercial planes. It would tend to make sense that the UAV model was most likely forked from there, at least initially.

Even with respect to radio interference in commercial planes, the public is subjected to more misinformation than information. How many planes have crashed due to cellphone use? Have there ever been any? Speaking specifically of commercial GPS use, was all the uproar about LightSquared just bullshit? They weren't on the GPS band, they were only near it. I recall lots of authorities confidently ordering us to just trust them, that LS was a dire threat to civil aviation. If LightSquared is a threat to UAVs too, our "enemies" might want to buy their old equipment off eBay. For that matter, if USA military radios are so robust, why do they have to reserve half the usable spectrum for their private pristine use? Wouldn't they be just as comfortable on the Wifi band?

There are reasons why civilians might be reluctant to simply trust the confident assurances of random anonymous defense contractors.

> There are reasons why civilians might be reluctant to simply trust the confident assurances of random anonymous defense contractors.

Which is absolutely your prerogative, and I don't blame you. You make some very good points. I would hope you might also be able to see my perspective: reading "facts" that are incorrect and not being able to add real value to the conversation. Sure, I chose my job, and the frustration comes with the territory.

> You've done an exceptional job of proving you have no idea what you're talking about. (Hint: I write software for a govt. contractor that builds drones and their control stations)

That's great, but you did not actually address the parent's point that GPS is jammable. Perhaps you are thinking of other methods of navigation that are accurate enough for long-term flying?

Enlighten us, please, rather than responding to people with your job description. Otherwise, you're not really adding anything to the conversation.

I can say that jamming GPS is a not a concern, hence my comment.

Celestial navigation can get you to within optical range of a target, so can inertial. Neither celestial nor inertial nav systems are jammable for cheap, if at all. Celestial navigation is well understood mature missile guidance tech.

Drones don't need GPS they can look at the ground in the day or the stars at night. (ED: Both of these are 20+ year old approaches.) Note, I don't work with this stuff so I can make that comment. Unfortunately, anyone who has specific knowledge is likely unable to respond legally.

Also, jamming is far harder than you might think; anything you use becomes a really easy target.

Edit: After some thought, as long as your internal clock is accurate and you can keep an accurate orientation as to what's 'up', the stars should be able to give you arbitrary lat/long position. Weather + barometer + location + temperate gives you fairly accurate altitude.

I don't think it's even a question of jamming the GPS network.

Once you start doing the math on "I'm going to keep 1,000+ nodes in a mesh network from operating a phased array capable of keeping a lock on the signal from a constellation of satellites", things like jamming become extremely non-trivial, and only operate over relatively short timeframes or with hugely vulnerable base stations.

Most people are thinking about the capabilities of a single drone -- that's really not the way to think about drone swarms. In much the way that it doesn't make sense to talk about the capabilities of cloud services like AWS in terms of the limitations of single computers or switches of hardware, but in terms of the ability of the system as a whole to emulate high(er) performance equipment.

Similarly, once we talk about real drone swarms acting as a locally networked, single functional unit, presenting itself as single emulated devices or functions that the drones carry out, the whole game changes from what's effective techniques against single, standalone drones.

So to recap, drone swarm's GPS downlink is probably a phased array or something like that distributed over the mesh network of drones (using the local radio gear as localized timing and positioning information between mesh nodes), and nowhere near as easy to jam as it is for a standalone drone (at least, for approximate location; then using cameras and local sensors for say, fine bomb placement).

tl;dr: The power of friendship works for drones too, not just comic book heroes.

As was said by irishcoffee upthread, jamming GPS is not a concern. If you think carefully about that statement for a minute or two, you'll probably figure out why: the incredibly weak, trivially-noise-jammed-into-oblivion GPS signal isn't used as a primary source of position data in battlefield conditions.

I expect that military GPS signals do have some crypto to help prevent spoofing, but that doesn't do any good when you can't hear the signal over the enemy's noise.

Good thing cruise missiles are self guided, then. Gyroscopes arent easy to jam.

For a few hundred dollars, someone will make the drones self aware with pre planned routings along optically recognizable waypoints. No more need for a comms channel.

What then?

Tomahawk missiles were doing exactly this and more back in the 1990's. Shoot off a few and they would communicate amongst themselves, detect (or get blown up by) threats, and alert the other Tomahawks in the swarm who would then reroute to avoid that threat. Additionally, the missiles shared weighted target lists and knew if a missile on the way to a target was destroyed or disabled, and so would reroute in-flight missiles -if needed- in order to hit the highest-priority targets.

That is still relying on comms which I pointed out will be solved eventually (if it hasn't already).

Tomahawks can operate fully autonomously with no substantial loss of functionality. Indeed, picture a swarm of one, and you're operating in this mode. The missile commlinks serve to enhance the missile's effectiveness; they are not the sole source of its effectiveness.

Here's something for you to puzzle over:

Your missile has up-to-date topographical maps of an area, and as-up-to-date-as-possible maps of all known anti-missile sites in the area. Thing is that you're not going to know about all of the anti-missile sites: many sites are portable (thus they move around), and some remain offline and nearly impossible to detect until enemy command determines that that site can kill a missile.

If your missile doesn't have commlinks, how do you tell it to avoid anti-missile sites discovered after the missile was launched?

Additionally, a careful reading of the comment that you replied to will reveal that the Tomahawk has -since the 1990's- been a fully autonomous missile fully capable of flying a pre-determined path to a target. Indeed, that is the least interesting part of the missile's capabilities. In short, we've had "self aware" warhead-tipped "drones" since at least the 1990's.

Thanks for all the insights about the Tomahawks. What OP is refering to is cheaper, smaller and therefore in much more numbers deployable fighting drones. Each and every anti-missile site, be it fixed or portable has it's limits on what they can defend against.

I also assume that OP was refering to drones that could go through current missile defence systems because they can't be easily identified, neither visually nor with radar.

Yes, we (or at least you guys in the US) had them in the form of expensive, capable systems, but not at a smaller, cheaper or whatnot scale.

Also, can't Tomahawk launches not be detected by a some signatures they emit? I assume that modern mil space radar systems are able to do that.

There are some ways to communicate that are significantly harder to jam than others, plus the degree of autonomy these systems have will only increase over time.

I'm surprised weaponized drones aren't displacing cheap rockets.

A big reason for the huge price tag and many delays has to do with the very broken military acquisition process. It's a mess, a huge problem that not many people have much insight into and fewer posses the real motivation and capability to change.

You can make a fairly valid parallel can be made with web designers and bad clients. You start off with a clear goal and reasonable costs and deliverables, and through iteration after iteration of needless revisions, terrible requests, and changed requirements you end up with a clown hat, a fart button, all in comic sans at 10x the original estimate.

---

Nobody wants to dogfight – the F-35 wasn't designed for it and modern strategy has nothing to do with it. You destroy the enemy with smart missiles long before either of you sees the other – F35 stealth is designed to delay detection as long as possible. This report takes the test completely out of context. It wasn't a wargame between A and B, it was a very specific situation under severe restrictions used to test the boundaries of the F-35's role. A real wargame would end with both sides never getting within miles. Not understanding the context of why the test was done leads to real misunderstanding.

----

As for drones - it is true they have a huge future, but they aren't the only solution. Modern air superiority battles are as much about stealth and electronic warfare as anything else. If your entire enemy is remote-controlled robots, your real target becomes the communications. There is very real value to having a human in the theater of battle – many of the strengths of the F-35 revolve around it as a platform for the interaction of many different communications, intelligence, and weapons assets ... the pilot is the command and control center for a whole network of activity which is much more resilient than remote control satellite RC planes and missiles.

Dog fighting only happens when absolutely everything else fails and both sides can bring each other into the stone age of point-and-shoot flying.

---

In a real war with a competent enemy, the first target (if we aren't all erased in a nuclear inferno) is going to be communications satellites.

Remember that a significant portion of how America fights battles is winning wars on paper so nobody ever thinks it's a good idea to start a real one, and by that I mean a war against a competent enemy playing at the same level – not regional conflicts and proxy wars like we've been fighting since WWII.

It's perfectly reasonable to say that in the future, human pilots won't be dogfighting, that's the job of missiles and drones, the human's job is to exercise tactical control close enough to not need satellite links.

But in that case, if the F-35's real job is to serve as a sort of fast, stealthy Awacs, then surely it shouldn't incur all the compromises inherent in building a plane that looks like it was supposed to replace the F-16?

It's hard to know these thing unless someone tries it, but probably not.

Air defences are difficult to overwhelm. Gaza/Hamas recently tried to overwhelm Israeli missile defence with primitive rockets with very no success. The rockets were very primitive but had the advantage of a short range, which is harder to defend against. OTOH, the intercepting missiles cost orders of magnitude more than the offensive rockets, so it might be possible to drain the ammunition supply.

"loiter over an enemy airspace" is only possible if you already have aerial control.

Overall, armies are conservative. It's dangerous to try and build a strategy around fighting styles that have never been tried at scale.

It's probably worth researching though.

> OTOH, the intercepting missiles cost orders of magnitude more than the offensive rockets, so it might be possible to drain the ammunition supply.

This is the key, I think. As long as you can pound your enemy with cheap hardware forcing him to use up his more expensive one, you're actually successful in your attack - you're bleeding out their economy.

Israel's rocket defense system, the one used to stop the attacks mentioned in the parent post is really good at being frugal in which rockets it shoots down. It doesn't just try to shoot down every rocket coming at it, it plots them and prioritizes ones that will most likely hit valuable or populated targets. It lets the ones that are going to land in empty fields go, since the risk and damage is minimal.

Pretty much... https://www.nsfwcorp.com/dispatch/cheap-drones/

A pragmatist would say that you are right. However ethics teaches us that you are wrong.

Remotely destroying things is not going to help us as a race. Keeping the human in the danger zone helps prevent violence, everybody has something to lose.

In practice, in recent battles, the US quickly gains air superiority and US fighter pilots face little if any challenge or danger. I'm not sure it's much of prevention.

Before manned US fighters enter a war zone, cruise missile attacks, HARM missiles, and stealth bombers usually take out the air defense system and bomb runways. The US really hasn't faced a challenge to air superiority in decades.

I'd prefer we find other ways to keep out of war rather than a few hundred pilots at negligible risk.

There is a reason that we fight our battles with guns and bombs rather than swords and shields; removing the soldier from danger insofar as is feasible while maintaining combat superiority has been the aim of most all warfare technology developments since the dawn of time.

While you can most certainly argue against the ethics of war, if you're going to permit war as a reality of human existence, then I'd say it's quite unethical to insist on waging said war with technologies that put your soldiers at any further risk than necessary.

It's not like we aim to fight fair wars.

"However ethics teaches us that you are wrong."

Well, there is nothing ethical about war. People in horses killed easily those not in horses. Then people in horses were killed easily with artillery as it improved.

The same happened with ships, making boarding unnecessary.

You know what they said about tanks when they appeared: they were not ethical. Then planes made possible to shoot someone kilometers away with heavy artillery(that before planes could not see far away). Not ethical at all.

Then came air bombers, the nuclear bomb. Today the US kills 10.000 people in Iraq annually with air drones just pushing buttons thousands of miles away.

It is already happening on one side. The only difference is when the other side starts using it too. Then we will consider it unethical(after decades using it on others).

When the machine gun was invented, Europeans congratulated themselves that they were the only ones with machine guns. After all, such a nasty weapon of mass death wouldn't be something that Europeans would ethically use on other Europeans.

That didn't last long.

On a micro level the opposite is true. The worst that's going to happen to a drone operator for not firing first is a reprimand for losing the equipment.

On a policymaking level, I can't think of too many air strikes called off from fear of the airmen's lives. Fear of terrorist reprisals, foreign policy debacles and ending up with people on the ground are far bigger issues to be considered which won't go away with drones.

Reply on news.ycombinator.com