Schneier on Security

Clive Robinson • June 4, 2025 7:36 AM

@ Bruce, ALL,

With regards the quote of,

“If the Ukrainians could sneak drones so close to major air bases in a police state such as Russia, what is to prevent the Chinese from doing the same with U.S. air bases? Or the Pakistanis with Indian air bases? Or the North Koreans with South Korean air bases? Militaries that thought they had secured their air bases with electrified fences and guard posts will now have to reckon with the threat from the skies posed by cheap, ubiquitous drones that cFan be easily modified for military use.”

This is nothing new.

It happened since before mankind could write, when he discovered that throwing a rock down on prey was effective as a first step in improving their chances of survival, either by putting food on the table as it were or eliminating a threat.

The military have always known,

“The advantage of the high ground.”

And as each technology advanced, they have exploited it.

You might as well look up the concept of the “Mass Driver” and similar Space Weapons. One is the notion of “rocks from space” where you simply take an asteroid or similar that’s outside of Earth’s gravity well and simply push it in on the right tragectory.

Currently we have no way to stop it, the only defence is the same as it is for drones,

1, Not have your position be identified.
2, Not be there when the weapons get there.”

This is in part what ICBMs IRBMs and now hypersonic cruise missiles are designed to built up areas and population centers.

It’s why I’ve mentioned in the past that the ten or so US carrier fleets are a waste of money militarily.

They can not hide nor are they comparatively mobile.

You can “invert the principle” with submersible weapons that can not be seen.

Think about a torpedo it’s a very basic submerged drone.

Now take it a step further, why bother sending nukes by express delivery ICBM at a great expense mark up, when you could build for a fraction of the price a bottom crawling nuclear mine…

It would just have to explode under water and as a result the ships could nolonger float. Or it could stand off outside a harbour that had appropriate land around it and the resulting explosion cause a tsunami type bore-race or tidelwave.

It’s why “infantry” are still current on battle fields they can hide and they can run.

As I’ve remarked in the past,

“There is no such thing as an accident, only lack of knowledge and the time to act on it.”

Which applies equally to offensive and defensive warfare regardless of the latest technology.

At the end of the day success depends on,

1, Denying your enemy knowledge.
2, Having time to react to your knowledge of the enemy.

That is the two essential fundamentals of warfare, and we’ve known it for at least four thousand years.

Clive Robinson • June 4, 2025 3:04 PM

@ Andrew Duane,

By all means

Just one condition if you ever meet Bruce, buy him two drinks, one for being our gracious host, and one for Bruce to pass on if I ever get to meet him.

Clive Robinson • June 5, 2025 12:17 AM

@ Moz,

A thought for you…

You say,

“Australia paying the USA hundreds of billions of dollars for the chance to maybe have a go in a US nuclear submarine one day.”

Something the US administration is going to dangle untill Australia is “bled white” (history shows this over and over).

The deal with France would in all probability have delivered that goal considerably quicker and a lot less expensively (remember I’m a UK person and the UK has it’s hand in Australia’s pocket on this US never never deal).

There is no reason at all why Australia could not develop it’s own vehicle based nuclear energy sources.

“… we’d need some way to deliver them since they don’t have the range to get from Australia to Vancouver by themselves.”

Consider two things,

1, Nuclear batteries last from 15-100 years based on half life of heat generating capacity.

2, The difference between a small nuclear reactor and a low yield nuclear device or considerable “dirty bomb” is not that great, if you know how to design them to be that way (mostly designers go the other way for “radiological safety reasons” and that’s one of the reasons nuclear powered rockets have still not happened after well over a half century, of which Project Pluto was perhaps the only publicly known working example).

See,

https://en.wikipedia.org/wiki/Project_Pluto

And the “curious Russian missile test” less than a decade ago when a test tripped radiation detectors thousands of miles away,

If you prefer a light hearted watch rather than a serious read,

https://m.youtube.com/watch?v=HPrFxfJ7WGw

The thing is the Project Pluto engine was a success and a “ram jet” is in reality little different to a jet engine with no moving parts. Further you don’t need the “working fluid” to be air, as you don’t need it for the O2 to burn hydrocarbons. Other “non ideal gases” will work.

Someone did some calculations and development work on doing this to “split water” to get hydrogen as a fuel…

So put a couple of things together, and as long as you are not Russia your chances of success are probably good…

Clive Robinson • June 5, 2025 6:41 AM

@ Moz, ResearcherZero,

With regards,

“did I mention we mine lithium here? And have a lot of salt beds if we want sodium”

And uranium, gold, rare earths, iron coal and vastly more including oil to the North West, etc, etc, etc.

You also have a very small population so mineral wealth per head is enormous, and although you have manufacturing etc Australia does not use very much of it’s natural resources. Worse Australia’s military forces etc are lets just say not adequate defence against the threats it faces (due to lack of warm body boot fillers and blood donors).

However the US under the present administration sees Australia as withholding what is “rightfully theirs” just as it does with other resource rich nations that are militarily weak.

Part of the reason is that the US whilst having about 34/800 of the Worlds population consumes around 1/2 of the worlds raw resources, and barely recycles any of them. So consumes 10-12 times the natural resources per head of population of the average for the rest of the world…

Thus the US administration sees it’s self as not having the resources it has a right to. Thus other nations

“Are holding it back”.

China also has a similar view which is why it gives aid to resource rich African nations in return for “shiny toys” and importing Chinese aligned technologists to keep the toys functioning. But certainly not training any of the population to become the required technologists so they can move out from under China. It’s a similar colonialism model to that which England practiced for hundreds of years to build an Empire to trade goods into.

On average the Australian population is better educated than the average Chinese “citizen”. So you have to ask why China are in effect “exporting” what is a significant chunk of their relatively few higher educated “citizens” that they have to under populated resource rich nations?.

Russia on the other hand is resource rich but it’s population nor subjugated others do not benefit and Russia does not reinvest what it earns via exports in the nation. Thus the long term plan is as it always has been for over a thousand years, which is to forcefully occupy and take from those prosperous nations around it as conquered colonial conquests and strip them of wealth for the benefit of the “brutish few” self appointed elite (of mostly criminals and their families these days, hence the term “Russian Mafia” has more than one meaning, and free flying lessons out of windows rather more common than else where).

So Russia, China, America are all building empires to feed resources and riches to the very few in their heartland at the expense of all others around them on the thuggish “might is right” principle.

Thus if you are seen to be exploitable then they will exploit you… The current US Administration is only giving a clear voice to long term US policy. Oh and it’s intent to destroy what it sees as it’s main rival… Which surprisingly to many is not Russia or China both of which it can exploit or even South America which it is exploiting much like China does with Africa, no it’s Europe it wants to destroy…

Australia is being seen as exploitable by China, Japan, Russia and the US and they are all actively trying to do so currently in various ways.

Clive Robinson • June 5, 2025 10:15 AM

@ Bob Paddock,

You note,

“It seems to me that such a wall of chains would foul the rotors of any Drone trying to enter?”

It’s why I said,

“And also unless troops on the ground are ready for it –and usually they are not at airfields– then fly other drones through the now left open hanger doors.”

I was thinking very similar thoughts, though more along the lines of just a wall of painted cardboard in a wooden frame on wheels or similar.

The problem I did not further think on, was “wind”… hanger doors are very high and very wide, thus even a gentle breeze would move things around because they would have a large “sail area” especially if they are very long as hanging things across even a narrow gap in the hanger doors would be.

Not an insoluble problem but as I’m not the one required to come up with “a practical physical solution” I went and had a cup of tea in preference 😉

Clive Robinson • June 6, 2025 12:09 PM

@ BillyBob, ALL,

With regards,

“So 10K small AI enabled suicide drones vs 10K soldiers. Drones might have upper hand since they are so nimble?”

The odds are not in the drones favour.

The Vietnam war produced statistics for all sorts of infantry weapons against the enemy.

The one most oft quoted is the 10,000:1 of bullets fired by US troops to deaths of those they were fighting.

The Vietnam war was the first to be “data processed by computer” to get battlefield statistics. What they showed was that killing experienced foot soldiers was incredibly difficult for both sides.

The advantage drones have is that they can get into defended positions like slit trenches that conventional weapons could not. However a drone can be stopped by very little so an attack with them needs certain types of planning by the attackers.

That is the first and second wave would be needed to “clear defenses” before subsequent waves came in.

Every form of defence has a weakness that can be exploited but it needs to be done in the right way.

For instance the Russians built “sheds” around their tanks and tracked guns. What the Ukrainians eventually ended up doing was flying in a “shaped charge” up against the side of the gun barrel. It did not damage the vehicle, but it killed the main armament. It’s actually quite hard to protect against such an attack as it was never considered in the original armoured vehicle design. Replacing a gun barrel is not a simple excercise, nor is making a main armament gun in the first place thus the tank / gun vehicle is effectively “off the field” for quite a period of time.

But consider infantry soldiers. There are three basic types of shooting they carry out,

1, Shoot to Scare
2, Shoot to wound
3, Shoot to kill

To maximise return against infantry the second option is best as this ties up not just the soldier but a couple of medics, stretcher bearers, doctors, nurses etc for quite some time.

When it comes to “red tab” Staff Officers especially in militaries like Russia runs. Then the third option is most effective because you “behead the viper” and the body just writhes uselessly.

The first option is used to make troops duck down in their slit trenches etc and you “sweep it across” coming behind are soldiers on your side who throw hand grenades and other shrapnel devices into the trenches to wound, maim, and kill the enemy troops.

It’s a way better technique than the “salt n pepper” attacks from WWI and WW2 that during the Falklands war got so many UK troops killed and wounded.

I must admit if I was attacking troops in trenches with drones I would consider “white phosphorus” magnesium and even PTF Flare material devices and similar. Even crude petrol bombs with added sugar is really not what you want coming into a trench burning up not just you but the oxygen you need to survive.

Remember a respiratory might keep out chemical agents but you still need oxygen to live. It’s the idea behind the so called hyperbaric weapons that were used agains bunkers in caves in Afghanistan.

Simple “Fuel Air Explosives”(FAE/FAX) are relatively trivial to construct, the thing is they not just send flames and a blast wave into every nook and cranny they suck the oxygen out as well.

Back in the late 1980’s we would use 5gallon containers about one third full of a mixture of petrol and other easily available fuels that stick and burn. We set them off using electrically detonated “Divers recalls”. The result of one hoisted up in a tree looked very like a nuclear detonation and there would be very little of the tree left. Worse they created quite a volumetric over pressure so you did not want to be close to one.

We got the idea from stories about “The Mob” in the US using a petrol can full of petrol and a hand grenade on a trip wire or similar strapped to it’s side. Also SOE training materials from WWII.

In open areas such devices are not very effective if you hit the ground face down and cover your head. However in an enclosed or semi enclosed space they are all to effective.

So the success of drones will not be decided by the fact they are “nimble” but by having an appropriate payload for the threat / target.

Clive Robinson • June 6, 2025 6:00 PM

@ not important,

With regards,

“How about using powerful laser pointers assembled as set with regulating focus to destroy drone optics?”

Over what sort of arc with respect to the drone?

The simple solution for drone flyers is to have some kind of “blinker” as they do on horses to restrict or block the field of view. Think going to “tunnel vision”.

And/Or have the camera on a rotating mount such that it can look from upwards to downwards as the operator thinks fit.

The drone camera has a quick forward view to see a target and approximate range and then looks down. The operator uses a basic inertial guidence system or just looks down to see “ground track” to fly closer to the target. Each time the drone gets half the distance closer it cuts the number of pointers in it’s forward field of view by four.

Also it could “ride the beam down” using a simple light detector system.

The use of a “stand off drone” that “paints the target” with a modulated laser at a very narrow IR light frequency range. And have the attack drone just ride in by looking for the “modulation signal” and pointing toward it.

These are all things that have been done with “smart bombs” in the past.

And I can think up several more tricks.

Speaking of which think of a stand off drone acting as a “sheep/guide dog” from “behind” it’s inertial-nav changes are used on a flock of attack drones in front so they in effect “fly ahead” in a “deaf, dumb and blind” mode “synced into the “guide dogs” movements. Then at a close point the attack drones switch modes from sync-flying to attack final run.

Whilst the number of variations of Attack(A) and Defence(D) are many they are not infinite. But the chance that a defender will have a system the attacker is vulnerable to is one over the number of “attack options” A to pick from multiplied by the number of “defence options” D so could be as low as 1/(A.D). The attack would however try to make the vulnerability window as narrow as possible whilst the defender would have to make each defence as broad as possible.

In the past this was done with RF jamming and “Low Probability of Intercept”(LPI) systems such as the various forms of “Spread Spectrum”. The jam power margin, was based on the attacker using as small a bandwidth as possible for data say 1kHz but jumped apparently randomly around channels in a 50MHz bandwidth. This ment to get equal power per Hz at the receiver the jammer needed to put out 50,000 times as much power when equidistant from the receiver as the attackers transmitter (adjust each by 1/(r^2) when not equidistant. So if the attacker is at half the range of the jammer the jammer signal would have to be four times as great as the jam equidistant jam margin.

The attacker could improve things by using “Forward Error Correction”(FEC) techniques so sending the data N times in a random time spread means the jammer if not in “fullband CW mode” would have to use N^2 times as much power.

The attacker could also put up a drone as a repeater with directional antennas thus like the guide dog keep it close to the attack drones and get that r^2 advantage over the jammer.

Oh and as for the jammer, it’s putting out so much power it’s going to look like a radar transmitter and thus it’s going to become a target it’s self to the equivalent of “Anti-Radiation Missiles”(ARM) or drones that ride it’s signal down and kill the jammer,

https://en.m.wikipedia.org/wiki/Anti-radiation_missile

This is why some years ago we got into the crazy world of EW, ECM, ECCM, ECCCM and so on…

The advantage usually belongs to a dynamic attacker in such scenarios. With the attack final run being a defenders only advantage if they can jam in that very very short time period.

Thus the trick as a defender, is not “static defence” but “active attack” to try to even the odds. Here the defender can also have the advantage of 1/(r^2) by having the jammer closer to the target than the attacker control transmitter. Unless the attacker can put the drone into some form of “auto-mode” where operator control for the final attack run is not required nor is external sensing.

This has been tried with “glide weapons” where the attacker brings the weapon to some small distance where the final run is in effect a fixed path. You then get into “circular error probable” calculations to decide how effective the glide weapon is or how big it’s warhead has to be,

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

As a rule of thumb the closer a weapon can be guided proportionately the smaller the payload has to be thus the smaller and less expensive the drone has to be…

Oh and we’ve moved on a bit… We now have coherent “Multiple Input Multiple Output”(MIMO) systems. Where the drone gets N signals from several transmitters that are coherent thus the effective information signal receive power goes up by N^2 at a selected point in space. Because coherent voltages add up to NV and effective power is (NV)^2.

There have been multiple books written on EW and ECM systems, I’ve a few in my “dead tree cave”. But they are never inexpensive and usually probability calculation heavy…

But a knowledge of simple “laws of nature” and geometry allows you to work much of it out for yourself from “basic principles”. Not hard, just tedious, but once you’ve “got it under your belt” as they say, you can usually linearize it[1]. And then “simply visualise” it without having to get a computer or calculator out.

[1] There is a standing joke rule of thumb in physics and engineering,

“Every thing is linear when plotted on log log paper using a thick enough marker pen.”

The thing is for many things it’s a “Close enough for government work” trueism. For my sins I still use a slide rule, because of it’s log scales allow just a fractional eye movement to calculate new values.

Clive Robinson • June 6, 2025 8:28 PM

@ lurker,

Looks like I’ve a use for the birthday book tokens I received 😉

Though when talking about “history” I immediately thought of Shakespeare and “Hoist by his own petard”.

Shakespeare was without any doubt a xontempory humourist of the masses. For such a line to go into one of his plays, says a lot about Elisabethen England and the common populations understanding of devices that much later Victorian’s called “Infernal machines”

But also “unmaned” does not imply sophisticated technology. Historically slow fused incendiary devices and later explosives were put on pack animals that would wander into town squares etc.

Even the CIA tried putting such things in cats, and during WWII small incendiary devices were put on bats in the hope they would roost in roofs and set them on fire.

B.F.Skinner of the “Skinner box” apparently tried to train pigeons to peck at buttons of japanese vessels in the hopes of making “guided weapons” that could not be jammed by wireless transmitters. Eventually the military went with spools of very thin but incredibly strong wire for both short range air to target missiles and torpedoes. Todays drones use lighter and longer fiber optics but the idea is the same.

As in life the wheel turns, and all that is new is the dirt that gets pushed aside as the rut gets longer through time.

People talk glibly of AI in weapons, in reality it’s not AI unless you mean the 1970’s / 80’s rules / decision tree systems.

Parts like image and voice recognition are more modern AI but they don’t as such “control the weapons” just act as sensors for basically rules based systems.

To many it’s sort of immaterial, because,

“AI in part is seen as AI in whole”

And this causes “muddled thinking” which others such as OpenAI, Microsoft etc ruthlessly exploit to their political etc advantage.