Schneier on Security

Clive Robinson • October 31, 2023 9:37 AM

@ Bruce,

Re : Power Projection.

“Soon it will be ships (probably with more expensive drones). Feels like this will be a major change in warfare.”

You left out the real advantage of drones over artillery and the like,

“They are lingering munitions that can be directed to an evasive even fast moving target”

Now apply the same logic to sea mines as drones with tactical nuclear war heads.

The US works abroad by “Projecting Power” with one or more of eleven carrier groups it expends multiple billions on by law.

On these are multiple aircraft groups from which “Power is projected”.

The Carrier thus defends the marine picket ships and they provide gunnary and similar to protect the carrier.

The carrier is vulnerable to three types of attack,

1, Air bourn such as missiles
2, Sub-sea such as submarines with torpedos or mines with tactical nukes
3, Space bourn projectile mass

We know from the goings on in the South China seas that both the Chinese and US are developing laser based defencive weapons to deal with even hypersonic missiles. We also kbow that the latest US carrier has more than twice the power generation that is required for current full-up running thus has been designed to have “new weapons” based on non chemical energy sources.

However as the Chinese know the best defence against carrier groups is to kill them before they get into their power projection range. Hence their hypersonic missile and cruise missile designs with upwards of 1000mile ranges.

Russia has similar ideas as can be seen by the number of launch pods carried on their larger systems. That are believed to be used in “punch-n-run” tactics, that is by quickly coming in close enough for their missiles to be in range but out of range of the enemy aircraft etc, fire a salvo and turn tail and either fire a second salvo or get well out of range and use the second salvo if pursuing vessels etc come in range.

Such missiles including hypersonic missiles are vulnerable to “directed energy weapons”.

Which is where sub-surface attacks come in, there is no suitable directed energy weapon that works under water we are still using technology people from WWII would recognise of sibmarines, torpedos and mines.

When it comes to “manned submarines” there is only so far you can go in making them stealthy whilst fast and they are very vulnerable close to or on the surface.

Thus taking people out and making smaller semi-autonomous systems is what torpedos is all about, whilst mines are a form of lingering munition. Combine the better elements of both and you end up with the equivalent of a bottom sitting mine, that can move to target, or an intercontinental torpedo. Armed with tactical nukes they can get in place from satellite and similar information from which a carrier group can not hide then launch an attack from inside the carrier groups defence perimiter.

https://www.usni.org/magazines/proceedings/2022/may/autonomous-nuclear-torpedoes-usher-dangerous-future

The UK were working on an interesting idea, of such a device but as a standoff weapon that was a torpedo and missile combined but was unlike the ship bourn US ASROC that drops the torpedo in by parachute.

The Chinese are are apparently going one further, using a boron based “zip fuel” in a ram jet system.

https://www.businessinsider.com/chinese-scientists-say-theyre-working-on-a-combination-missile-torpedo-2022-10?op=1

But these are not subsea “lingering munitions” but they could well be so with not as much design effort. Apparently Russia is working towards that,

https://nationalinterest.org/blog/reboot/russia’s-status-6-torpedo-true-nuclear-doomsday-weapon-172360

But the third item on the list is the one that should worry many people. Whilst man made missiles can be in theory disabled / destroyed by directed energy weapons, the “rocks from space” or “mass driver” idea is not. In theory it’s simple you go and take mass already in space and direct it at high velocity towards your target. The energy released is not that difficult to work out nor are most other asspects. Much of it is already public knowledge, due to the fact that we know rocks from space are a real existential threat, and people are looking at ways to deal with them.

But it appears the US atleast has decided that one way to defend it’s carrier groups is to take out satellites from space, thereby try to stop them being seen.

https://www.nationaldefensemagazine.org/articles/2021/4/15/seizure-of-satellites-could-serve-as-option-to-defend-us-spacecraft

Clive Robinson • October 31, 2023 11:38 AM

@ wiredog,

“Mass drivers would first have to be lifted into orbit”

The plans defending against them back in the 1980’s was that an attacket would use mass that was already up there, that is “rocks from space” already moving around the solar system, that would be “waste” from an assumed up and running asteroid mining by 2030. But also before that any sizable old satelites and any other pile of space junk that could be amalgamated to survive reentry.

Obviously nearly fifty years later we don’t have asteroid mining due to political economic reasons. However it does appear to be an idea that is getting funding again which brings back the question of what to do with the ~80% waste mass in the future.

Asteroids are not exactly in “Stab O” but have settled into various orbits. But if we start moving them around, we realy don’t know what will happen. But also the qiestion of who owns them and despute resolution that could turn into conflict it’s self,

https://www.bbc.com/news/technology-61421787

So yes wilst mass drivers are not an immediate threat some consider them something to work on as far as defence systems go. Because stopping a large rock aimed at a major city does not give you many options, especially if you can not stop a rock only ten meters across heading for a patch of ocean etc.

“Back in the 60’s people were discussing the idea of Project Thor, AKA “Rods From God”, which puts loitering munitions in orbit. Again, though, the cost of getting them there is high and they are fairly easy to detect and track.”

Not just the 1960’s when space flight genuinely was expensive.

These days the cost per metric ton lifted has dropped significantly when compared to other economic measures. And with individuals being able to set up Corps that can build rockets and launch high tonnage payloads out of Earth orbit, I can understand why defense planners have got it on the “to do” or “research now” lists.

You and I are of an age where we probably won’t see such worries, but our children probably will. Space will get industrialized, or the human race will not survive as it currently is, that’s a reasonable certainty.

The problem as always is we never properly take care of the waste, even when we know it’s highly dangerous (think nuclear and chemical industries).

History said that “alcohol was the peoples curse”, then others pointed their fingers at religion and now politics. The fact is we have a bad habit of being both tribal and violent when resources are involved. History gives us examples such as “Water Wars” and “salted earth” and similar. When you cut down through the nonsense most European wars were about the control of resources.

The history of “gun boat diplomacy” tells you much about “Projecting Power” and that infomous saying of “bomb them back to the stone ages” just a more modern version… As was once observed,

“Weapons are made for war, profit, and subjugation it’s the way Empires are built.”

And as various people have thought “Why be a king of a country, when you can be Emperor of a continent”…

Clive Robinson • November 1, 2023 3:04 AM

@ Newt, ALL,

Re : Drones for ElInt.

“Even further musing – if a swarm of UAVs is locally networked, it would only need a couple members to be high or far enough from the source of the jamming to pull in accurate geo-location.”

As I’ve posted before, it does not require more than a couple of drones and they don’t need to be networked together at the drones, thus if one gets shot down it’s immediately replaced by another as there is no interdependence.

But I also provided the calculations that you might find interesting.

Take an ordinary cheap and nasty VHF/UHF handset like the UV5 from quite a number of Chinese manufacturers for around 20USD. Which we know the Russian Troops were using at one point, because due to corruption in their armed forces some general etc did not buy the ~5000USD encrypted military radios he had been given the money for he just pocketed it and payed a few people off.

Well those UV5 radios being near enough “line of sight” have maybe a mile range soldier to soldier when soldiers are down on their knees, rather less if they are on their belly or in a trench. However they radiate upwards just as well and can be easily picked up in space at the above LEO of the ISS (I’ve done it so have many others) so a thousand or more miles of range fairly easily.

Now if we assume a hobby drone that can lift a thumb stick USB “Software Defined Radio” connected to a “Gumstick SBC” which is connected to a tiny USB WiFi dongle with a battery good for an hour or so, you have a quite effective remote ElInt Sensor. Hand making antennas to have good back to front ratio makes tthe SDR much more sensitive to the enemy emmissions, but also significantly reduces the WiFi emmissions towards the enemy, whilst incrrasing it to the “ground crew” operators who can likewise use a very high gain antenna pointed at the drone thus significantly reducing their “visability” to the enemy.

But what is the “sensor range” well assuming the drone flies at 300meters up –which quite a few can– then it’s radio horizon is upto 60miles.

At that distance Russian weapons radar would not be able to see the drone especially if it moved quite slowly, but it might be seen if a poorly designed front facing antenna re-radiates (much like the half wave strips of foil called “chaff” that jamed German Radar during WWII). Solving the re-radiation issue is not particularly difficult but tends to reduce reception band width.

So sending over a very cheap sacrificial drone hanging halfwaves to act like chaff for the Russian systems would be like the old WWII technique of “gardening”. Where a plane would fly into German air space near a German docks to cause the Germans to send out an identical warning message on all channels, thus “light themselves up” for British inteligence at Bletchly Park.

As you note,

“And if anything were broadcasting such a powerful/noisy signal for jamming a large area, it seems like that would be self-nomination as target numero uno.”

As many of those drone jammers were designed for civilian use for places like airports and conference centers and the like, they are sort of designed to be held and used like a shotgun… After a couple of “Nowt left but ‘is boots” finding a volunteer to use one is going to be a “pistol upside his head” type inducement from a more senior NCO thug which are unfortunately ten a penny in the Russian Army.

There are other systems available but they are very far from being inexpensive. But even though designed for “static deployment” they could be mounted on an armoured vehicle…

https://m.youtube.com/watch?v=LCMbHbtMq8k

But as the article @Bruce links to notes if a 400USD drone can take out a high end main battle tank… What chance the armoured vehicle, that can not even assuming ideal circumstances do a quater the speed of the drone, which is also many times more maneuverable as well…

Clive Robinson • November 1, 2023 10:32 AM

@ Winter,

Yeh I’m aware of the basic methods, back some years ago now I worked on a “radar altimiter” system that used four microwave beams to measure not just contours but doppler effects as well the spec called for it to be used at less than ~1300ft / 400m from a 30cm diameter body… So I think you can guess where that was probably headed.

As for star mapping yup I did a prototype, you take yourself a 3U cube sat (30x10x10cm) with power bus solar cells and NiCad batteries and put in the midle down the axial length a striped down “Consumer Off The Shelf”(COTS) large sensor SLR body with the largest telephoto it can hold. And the equivalent of a modern web cam or smart phone camera looking in exactly the opposit direction. The idea, take a photo from each at the same moment and then “Oh so slowely” send them down at 9600baud – 384kbps in the low end of the UHF band… Then use a ground based computer to do the starfield mapping… It was for an “educational cube sat” but… It turned out that back then a modern SLR breached US DoD civilian resolution limits for LEO so the idea got canned.

But as I’ve noted you can ride all sorts of contours if you go avout it the right way, including air preasure. Which actually surprised me as I thought it was just not stable enough in even short time periods. Turns out though you can actually predict it sufficiently well in near real time to make it viable enough for crude “get me to the ball park” navigation. Thus saving a lot of energy and being somewhat stealthy as you don’t have to emit a signal.

Clive Robinson • November 3, 2023 5:21 PM

@ emily’s post, ALL,

Re : Small can hide and skip.

“How about massed phased array jamming and laser targeting and destruction against these relatively slow moving objects?”

So,

The first question is

“jam what?”

I covered jamming basics just a few hours ago,

https://www.schneier.com/blog/archives/2023/10/friday-squid-blogging-on-the-ugliness-of-squid-fishing.html/#comment-428308

But to amplify, most drone jamming equipment only works on a tiny handfull of frequencies that are used “commercially” due to regulation. National Security exemptions say you can ignore the regulations thus any of very very many frequencies can be used by your opponent. If you don’t know what frequencies are used then you can not specifically jam them. So unless you have very good ElInt your only resort is very wide band jamming, which really is a last desperate option.

But importantly remember that wide band jamming effects your troops rather more than it does your opponent. Because the opponent starts further away and may well have a terrain advantage that you don’t thus gets way more “safe range” than you do.

That is if you are using hand-held jammers or backpack jammers imagine you are in trenches in a gully or other low,lying area with rocky or other RF-Blocking terain close to you. Your opponet simply flies iat you “in the blind” like walking up to some one by “staying in the shadows”. Even worse the oponent could be “looking down” from high terrain, another drone, aircraft, or even a satellite using ultra high or microwave frequencies. These small attack drones can fly within inches of the ground and thus use any cover. The fact that your opponent has a significant lookdown advantage can be further added to by using quite directional antennas. Think a small mesh, chromed plastic, or even baco-foil on cardboard “horn antenna” looking out the back of the drone. It can give the opponent upto a 30dB front to back ratio advantage, which means you need a thousand times the RF Power… Even a VHF antenna set up to give a deep cardiod null can give upto 40dB null depth which means 10,000 times the power.

Think what you would get if rather than using RF they used Infra-red diodes in funnels on the drone and class III or above lasers?

But even if they are using RF… Jamming only works if,

1, The receiver is on AND
2, The drone can not function with out a control signal.

Neither is true for even commercial drones of any size…

Because changes are just upgradable software patches/overwrites anyway, and the skills to do this are not as unique as many would like to think.

Also remember that whilst you can jam a receiver, you can not jam a transmitter… That is the video signal from the drone radiates out no matter what you do. There are easy tricks the drone opperator can do using phased antennas to make a synthesised antenna of such a narrow beamwidth as long as their receiver front end does not get overloaded, then the signal will get through.

Without going into the messy details the operator can have a receiver that has upto 90dB dynamic range and a sensitivity of -110dB relitively inexpensively. Attached to an antenna giving 13dB of gain to the drone and 30dB of rejection to the jammer so effectively a 43dB advantage. Which says as the jammer operator you have to overcome that margin as a minimum which means you need 20,000 times power advantage… which is why military jammers tend to have outputs in the multiple thousands of watts.

Which makes them like a light house on a moonless night, a “very easy thing to spot from miles away”.

I won’t say how to do it, as it’s easy to look up and it’s not difficult to do, but you can make a drone “fly down the beam” of the jammer… Which if it’s one of those commercial ones you use like holding a shotgun … The person holding it is going to have to be “out of cover” with a high chance they are going to turn into smoking boots and not much else.

Thus jammers are very easy to hit at the antenna as “They paint themselves to missiles” and drones alike. The technology is well documented and taught to undergrads and it’s been in “anti-radar missiles for over half a century.

If you have a Google you can find the name of a rebel leader who’s low power satellite phone had such a missile “fly down the beam” in the 1990’s.

With it having been alledged that his demise was what persuaded Osama bin Laden to stop using any kind of radio system. Instead switching to using human couriers who stuck messages hidden in gay and other porn vidios on thumb drives up their back passages to crall out across the Pakistani boarder amoungst others. (stories of which were apparently very popular in US right-wing circles as the failed missile strike linked Bill Clinton’s sex scandle to both the escape of Osama and loss of SigInt inteligence).

But as some have pointed out the time lines given by US right wing Politico’s over Osama bin Laden turning his Imarsat phone (873682505331) off don’t match either,

https://slate.com/news-and-politics/2005/12/don-t-blame-the-times-for-the-satphone-leak.html

Actually the evidence is more alobg the lines that after the failed US missle attack, Osama stopped using that phone and started to use a different system with voice scramblers upto and shortly after 9/11 and it was only after that his taste in phones changed.

Jamming is a high risk and high cost fairly futile endevor without good electronic intelligence…

Similar reasoning applies to “Directed Energy” weapons like lasers. And whilst in the scheme of flying objects drones are slow at upto 160kph unlike other aircraft and missiles that fly faster they are very maneuverable.

Clive Robinson • November 4, 2023 5:31 AM

@ emily’s post, ALL

Re : Difference between jamming and radar and being still and stealthy.

“What about drone vs. drone ?”

It falls first to the vexed question of “bringing on target” you can not acurately locate and engage what you can not sense.

That is you have to detect the enemy drone at a safe distane for the hight it is flying at[1]. To do this you need some sort of energy signiture off of the drone[2] either by it emitting it or it reflecting off of it sufficiently at your detectors.

US war doctrine in recent times has been to send in stealth aircraft with anti-radar missiles to take out defence missile systems then send in more conventional aircraft. However stealth tech has a number of Achilles heels which people are designing detection systems to[3].

Drones however have a stealth advantage that always works over detection systems, in that they have a very small cross sectional area and don’t use a lot of power (however some drones do have fast rotating reflective surfaces called propellers that give spikes in the return frequency spectrum that can be “dug out” with DSP and now AI).

Which means with drones the defender has to either accept a significantly shorter detection range or significantly up the power of their EM emitters.

Thus the same issues that effect jammers apply to all EM emitters including Radar and Lidar.

But as noted earlier, drones can fly very low thus hide in the shadows behind objects through which line of sight emitters can not go. The only defence to this is a “high picket fence” system where you have several high altitude pickets looking down thus minimising the shadows. You see the “big boy” version of this play out with naval “Power Projection” by “carrier groups” but as war planners are now having the heebie-geebies about, the picket system fails badly to surface resting loitering munitions with some degree of Signal Processing AI automation.

A drone powered down just sitting on the ground is incredably difficult to see and with a little camouflage has almost the same detection issues as land-mines. However AI drones have the advantage over land mines in that they can move. So they can set down infront of an advancing column and wait untill it is unknowingly upon them then they spring up and down on targets so fast that they get right inside the columns defence loop.

From a non battle involved observers position it makes for interesting planning strategies.

But the take away is drones are already a century after the tank a military planners nightmare of a larger calibre.

With worse a lot worse yet to come with improvments in “loitering munition” techniques just starting to show it’s power. It’s not hard to see how the addition of 1980’s level Signal Processing and AI you can now get out of a 2USD SoC will make most current battle plans redundant as internal inertial navigation by MEMS devices will be sufficient over short ranges thus loitering munitions make jamming / radar systems very inadvisable as a defensive measure.

It will devolve into a game of who can get inside of the others “OODA Loop”[4].

Drones may not have much in the way of warhead mass capability, but they are very difficult to spot, harder to stop and only a fraction of the capabilities they will have within a couple of years, and even now cost less than a month of labourers / soldiers wages… Which is a lot less than a non-smart artillery shell. Interestingly as the Ukranian’s are showing they can be made in garages and even be shipped on push bikes.

But consider the other issue, drones don’t fly in “balistic curves” nor do they have to follow other predictable paths so “Iron Dome” style defence systems will fail to intercept them if drones implement even simple path randomisation. Not something various people would want widely known as it would weaken their position in the eyes of those they are supposadly defending.

[1] Depending on the type of drone and how a second drone might disable it, you end up with a “balistic flight” or “decaying orbit” depending on which view point you take. The 1.5kg munition it is carrying on disabling the drone is most likely to follow such a curve and travel the furthest balistically. Obviously the higher the drone the greater that distance. Thus the optimal stratagy for a drone pilot might be a “stand off” “lob shot” as developed for pilots to survive droping tactical nuclear ordinance from fighter-bombers.

[2] It turns out that making a stealthy vehicle is a question of “radar profiles” which is a “shapes and materials” issue. Make it with tubes of aluminium and it will have an all round reflection of high order to viewed surface area (think “disco mirror” ball). Make it instead from a triangular tube of carbon fiber and it does not have an all round reflection and as carbon fiber partially “absorbs” RF and turns it into heat as well as scatter reflecting some, and if made thin enough or of slotted surface will even alow most of the RF energy to pass through. Thus even large wing surfaces if made of a sandwich of very thin carbon fiber over both sides of a carbon loaded foam will have high structural strength but very little radar reflective surface. Back in the 1970’s what electronic engineers jokingly called “Hundred ohm foam” which was a conductive carbon granual matrix in a foam designed to store “Dual In Line”(DIL) pin packaged chips safely against static electricity was found to have all sorts of other uses, like grip preasure sensors on robots, very low frequency vibration detectors and due to it’s EM absorbtion abilities a way to beat both doplar microwave and Infra-red sensors used in then expensive high end burglar alarms, and when calibrated effective antenuators in microwave wave guide and audio and RF anechoic chambers.

[3] What is called “Stealth” was about,

“Not reflecting directed energy back towards the emitter.”

And that is all about “shapes and materials” whilst a spherical mirror reflects back from every direction it’s actually quite a small area that reflects at 180 degrees. A flat surface broad side on reflects a lot more, and getting on for best of all is the tri-corner reflector used on sail boats and the like for being radar visable, and if you look close enough at push bike reflectors you see hundreds of very small inside corners across the inside surface. All materials have a frequency spectrum where they either transmit through (transparent) or absorbe radient energy the inverse of which is the frequency spectrum their surface can reflect. These effects can be combined constructively or destructively to reflecting energy back to an emitter, reflectors on boats and bikes are designed to be as constructively reflecting as possible for safety, stealth is designed to be as destructively as possible to avoid detection. Thus for avoiding 180 degree reflection you want no curved surfaces, no flat surfaces at 90degrees no internal angles at 90degrees this tends to make for the aero dynamics of a brick… Thus you end up with needing a lot of power to keep it in the air or accept some other compromises or both. High engine power means a lot of infra-red comes out the tail pipe and significantly limits range. But also those flat surfaces still reflect even if it is not directly at the emitter.

The IR emissions mean that heat seeking missiles and other thermal imaging systems have a much easier time, and the flat surfaces, can like a mirror flash very brilliantly to “offset radar” where the receiver is not colocated to the emitter.

[4] The “Observe, Orient, Decide, Act”(OODA) loop was thought up and developed as a fighting stratagy by USAF Colonel John Boyd in the 1960’s. The essentiall observation about it is “If your loop time is less than your opponents then over a enough engagments to average out random events you will win by having more players still on the board”.

en.wikipedia.org/wiki/OODA_loop

With drones being cranked out like sausages from a machine, and requiring no time to train unlike humans, OODA and reduced costs is very much where the next series of wars are going to be. With internal guidence and loitering and passive sensors with DSP and AI making the OOD phases invisable to the defender who has to do their entire OODA in less time than the loitering munition drones A phase… This means the defender very much has to improve the sensitivity and discrimination of their initial O phase –the only phase they can have an advantage in– and significantly reduce the time of the ODA phases, otherwise they are “dead meat”.