Comment by Tuna-Fish | Hacker News Reader

Comment by Tuna-Fish | Hacker News Reader

Parent Full thread

> Dogfighting was supposed to be archaic and improbable in 1968 as well, since who was going to get close and maneuver when you had guided missiles? Then the losses mounted, and they created TOPGUN.

There were some actually good reasons to believe that dogfighting was a part of a bygone era. When those planes used in '68 were being developed, missiles truly had the kind of superiority that their designers expected them to have. Then better countermeasures and anti-missile tactics were developed, and suddenly dogfighting was relevant again.

The first part is again true today. Missiles are easier to update than aircraft. Taking an air-to-air missile from the 90's and ripping out the laughably outdated electronics and replacing them with a modern CPU and targeting software makes it able to completely bypass all countermeasures available to current fighters. Right now, if an enemy gets to lock a missile at you from within the envelope where you can't simply outrun the missile until it's engine runs out, you might as well immediately eject. Your evasive maneuvers are not going to beat a missile that is getting mid-course updates from it's firing aircraft, which can tell your decoys apart from your plane, and which can plan and simulate the optimal attack vector in ways that the 90's missile designers could only dream of. Near future missile designs using much improved engines like ramjets or the throttleable ducted rocket with a secondary ram combustion stage on the MBDA Meteor seek to increase the engine performance and longevity of missiles to the point that you really can't ever even outrun them.

So currently no aircraft can remain flying with missiles in the air. There is, of course, always the possibility that we develop new countermeasures that can beat modern missiles, returning dogfighting to relevance. However, even if missiles always beat aircraft in the future, that doesn't make a F-35 a good fighter. Because, if missiles always beat planes, why not just build lots of cheap planes with a lot of missiles?

I'll admit I've been out of the field for a long time... I sort of cut ties and haven't read up on the latest tech in probably 15 years. Your point about missiles receiving mid-course updates is very interesting, though. I have a couple of questions...

Has any (unclassified) work been done on rear-mounted weaponry? I mean, if missiles can have modern CPUs and targeting software, can't the countermeasures have those as well? It seems like accurate, rear-mounted, auto-targeting machine guns (or heck, lasers for all I know) could still be an effective countermeasure, especially if you're in the position of trying to outrun one. That is to say, rather than trying to outmaneuver, outrunning would take you on a straight path, which would put the missile on a more-or-less straight path behind you until it hits. It seems like this is the perfect scenario for auto-targeting defensive cannons.

I have seen the Trophy countermeasure system for tanks work in a similar way, its not ridiculous to think it could be applied to planes.

Upon further review it looks like Raytheon is also making one (again for tanks.)

https://en.wikipedia.org/wiki/Trophy_%28countermeasure%29 https://en.wikipedia.org/wiki/Quick_Kill_active_protection_s...

Lasers have a big power to mass+volume problem for mounting them to fighters. Check out the anti-mortar laser systems for an idea of the size of current laser systems required to shoot down mortars. That's probably a good lower bound on the power required for an active A2A missile defense system. Similarly with using kinetic defense systems mounting a system with a high enough rate of fire along with the tracking and pointing mechanisms just doesn't work on a small plane like a fighter, we have systems like that for ships but they're huge.

Maybe, but at the cost of how much added weight? And even if the thing is approximately 100% effective I'd still think pilots would be very reluctant to voluntarily become a perfect target.

Your comments about modern air-to-air missiles may be correct, but in 1968, the beyond-visual-range (BVR) engagement was more of a bomber general's fantasy than a reality. Kill probability of the AIM-7 Sparrow missile in those days was less than 10% [1].

Pilots' accounts of air combat in Vietnam routinely mention the firing of a salvo of missiles --rather than just one-- when a firing solution presented itself, due to the notorious unreliability of missiles.

[1] https://en.wikipedia.org/wiki/AIM-7_Sparrow

The specific aircraft I was referring to, the F4 Phantom II, first flew in 1958 and was introduced to service in 1960. The missiles that the designers had in mind were the Sidewinder and the marketing promises of the Sparrow.

And also, the sparrow had a decent hit probability -- that is, unless the target aircraft released chaff, in which case it was very unlikely to hit.

So why don't ground-to-air missiles make all fighters obsolete?

(If the answer is that they lack "mid-course updates from it's firing aircraft," that sounds like something that could be jammed.)

Communication between plane and rocket has several options, one of which is effectively a laser pointer in the aircraft pointed at one in the rocket modulated with the information to be transmitted.

It's fast, lots of bandwidth, and it's totally unjammable. Or even detectable for that matter.

Unjammable by clouds? Wouldn't going into clouds be a way to evade in this case?

Depends on the frequency. Clouds are famously transparent for ultraviolet for instance. The sun floods the sky with it of course, but there's plenty of choice. But at the ranges we're talking about 10-20, maybe 100km ... yeah pretty much unjammable by clouds. Sufficiently intense rain or something like a snow storm would probably work though.

I doubt you'd be flying in the conditions that would block a laser from hitting it's target. In practice, there really aren't that many ways to "jam" a rocket, radio jamming really can't hit all the bands, and frequency-hopping SDR is very, very cheap now.

To my mind your best hope would be an EMP or a small missile-tracking gun.

> frequency-hopping SDR is very, very cheap now.

Which also makes broadband frequency-hopping jammers similarly cheap.

Frequency hopping doesn't make jammers more effective. So long as you cannot prevent the pattern of the transmitter, your only option is to jam all frequencies, which means you have a massive disadvantage in the power you need to emit, to the point that simple directional antennas probably make your task flatly impossible.

Laser communications are being studied because they are impossible to detect, not because the existing communications are too easy to jam.

(late edit:)

*predict

Because ground-to-air missiles need to be put on the ground before the engagement, and that's not always possible, I guess.

AA missiles start with a huge energy disadvantage. They start with a velocity and altitude of 0 so they need an engine powerful enough to catch the fighter and enough fuel to reach it's altitude. More power means more weight, more weight means more fuel, that equation results in a huge missile with minimal maneuverability.

No maneuverability = no hit. Surface to air missiles are intended to target low, slow, or bulky targets and fighters are outside those categories.

Sounds like what's needed is a bunch of high altitude blimps or cargo planes with air-to-air missiles.

And enough countermeasures they don't get picked off by ground-to-air missiles.

Or two-stage AA missiles with large "get it up in the air and fast ASAP" 1st stages and smaller "aim at the baddy" 2nd stages?

I highly encourage everybody to read the book 'Sidewinder: Creative Missile Development'. Missiles are not magical things that are just better; they had lots of issues and it took a radical US team to produce something really good. The Russians subsequently copied it, improved it and celebrated the birthday of the father of the Sidewinder!

Because stealth and sensors still play a huge role in your effective distance of engagement, don't they?

targeting hasn't improved as much as you think (hell, 1920s mechanical computers have about as much accuracy as a modern day cpu, so you can tick more in a second, so what?). there are still so many reasons for a missile to miss its target that even todays missiles are unreliable against an advanced craft with the right support.

thats the real reason they believe dog fighting is dead. having radar, drone, and satellite control of the sky is far above and beyond important. who cares if your multi core cpu enabled missle has a .0001% more accuracy when you know when the enemy's plain is fueling up in the hangar?

the US doesn't believe its going to get into a fight with a peer competitor, that's why they are ill concerned with dog fighting abilities.

>the US doesn't believe its going to get into a fight with a peer competitor, that's why they are ill concerned with dog fighting abilities.

Then why don't they buy more AC-130?

I'm not saying they are right in that assumption, and I'm not sure why you would suggest the AC-130.

they are more focused on drones for better or for worse

The focus of your previous post was that electronic gizmos make dogfighting obsolete and unnecessary. You can put far more electronics and weapons in an AC-130 than in an F-35. I'm only half-joking here. If you have no maneuverability or power advantage, then just forget about dogfighting altogether and take more weapons and countermeasures.

If they field this thing and try to rely on it, someone will build a purpose-built F-35 killer for a fraction of the cost and field 10x of them against it in battle; assuming anyone ever wants to have an air war with the US.

>they are more focused on drones for better or for worse

I agree drones and remotely piloted will eventually supplant piloted craft, but I think the Air Force is a generation of leadership away from that change.

> assuming anyone ever wants to have an air war with the US.

If the past serves as a guide, it's usually the other way around.

Because the AC-130 is far more vulnerable to MANPADs and so has had periods even in Iraq where it was restricted to night flying only due to losses.

> targeting hasn't improved as much as you think (hell, 1920s mechanical computers have about as much accuracy as a modern day cpu, so you can tick more in a second, so what?)

The primary advances are the ability to distinguish signals returned by chaff/flares from the real aircraft, and to have a model of the target aircaft's capabilities, so that on each targeting step the missile simulates the most extreme maneuvering it's capable in all directions, aiming the missile so that it is capable of keeping it's seeker on it and following regardless of how the plane flies. That's something that a higher tick rate is quite useful for.

(I'm agreeing with the parent post, just elaborating.)

Also, much like the weather, sometimes you can only make accurate predictions if you make a small prediction and correct it once you get feedback.

Much like you can get insanely high correlations on your predictions about the weather if you never predict more than 15 minutes out, the problem of predicting the other plane's response only become tractable when you can predict and correct over very short windows.

The higher clock rate doesn't just do more of the same faster, but rather, enable a sort of "phase shift", where the speed enables a new form of computation to work in a meaningful way.

Doesn't that imply that anti-missile missiles are just around the corner?

> Doesn't that imply that anti-missile missiles are just around the corner?

Anti-missile missiles already exist (missiles designed to intercept air-to-air missiles are still a ways off.)

Reply on news.ycombinator.com