Schneier on Security

Clive Robinson • August 16, 2005 11:56 AM

Just a thought,

After all the bally-ho about Iraq and WMD the only two things that could be brought against Iraq where,

1, A field rocket, could when it did not have a warhead, and with specialy selected motors and fuel, excead the limits placed by the UN… (100Km if I remember correctly).

2, They where developing unmand reconasance aircraft UAV’s to patrol the Iraq/Iran and other boarders.

Although in no way demonstrated it was considered that both systems might possibly carry a small munition of a chemical or biological nature.

So if the US goes to war over Iraq just developing (not deploying) UAV’s then they must in consiquence accept that other people (Mexico / Canada) might regard this as being an act of overt warfare…

Clive Robinson • August 17, 2005 9:56 AM

@Dark Fire

With regard to your OT post,

First off more than half the cammeras on the London underground do not work at any one time (due to water ingress in cables, vandalisum etc). Likewise a lot are actually not recorded any way (tapes to expensive to replace / not checked / not changed in recorders etc).
Few of the operators actually look at the CCTV screans except when other people are present. There was a report about this that made it’s way out into the public domain, unfortunatly few newspapers picked up on it at the time. It was suspected that it was leaked by a member of a Union who at the time where complaining vociverously at the time about the number of attacks on platform staff / cleaners. Transport For London (TfL) made noises, but as usual the organisations who day to day run the London Underground decided it would all go away, and guess what it has.

@Others
On advantage of the DHS buying UAV’s is that the price will drop which will make them more atractive to other organisations, and the price will keep dropping, untill they are at or below the price of a new helicopter…

With regard to the accident rate, the little information I have seen in Military type journals indicate, that there are four main areas where there are problems,

1, Take off / Landing (high risk)
2, Air Colision (low Risk)
3, Maintanence issues (low-medium risk)
4, Anti aircraft measures (high risk)

The last is not expected so much on a boarder (unless it’s an active battle field). Also the second is not likley to be an issue either in general you cannot fly along a countries land boarders it’s reserved air space. As for maintanence, as more UAV’s are manufactured, this will become less of an issue, as the part reliability will increase (think of 1950/60’s jet fighter aircraft reliability/crashes and compare to todays jet fighters).

The first issue is due to lack of training procedures on the UAV operators, the type of UAV’s (batle field designs) and environmental factors, also in early military systems and drones they where like ordinary RC models only larger, sort of designed to crash land..

Even for current small systems costing just a few thousand dolars (used for map making etc) it is getting to be a past issue.

One of the (minor) reasons NASA (amongst others) have been looking into solar powered high altitude aircraft is it kind of removes the first two issues at a stroke, as they can stay up almost indefinatly. There is for instance no reason why a radio isotope power generator (as used in deep space probes) cannot be used in an aircraft that is basically a glider, if this is also augmented by solar pannels on the wing surfaces. Like unmaned nuclear subs (yes there where prototypes made) they could be deployed almost indefinatly.

This would obviously not work for military UAV’s they are designed to be fast and manoverable to avoid being shot down, they often use small high performance engines (jet or infernal combustion) which are by their very nature cutting edge, therfore prone to failiure. Non military UAV’s of the high altitude type are going to be extreamly cheap to make, use off the shelf high reliability / redundant components. The only expensive bit is going to be the avionics / auto pilot system, and the price of that is going to drop like a brick in free fall as the market opens up (it is realy only a PC GPS and RC actuators with moderatly clever software).

Likewise for boarder type activities, you are not going to require the complex on board data procesing and signal processing systems or the covert satcoms systems. These are needed in military systems to reduce their probability of detection.

Another aspect of high altitude UAVs that are designed to remain aloft for long periods of time is the savings in manpower. If they are essentialy flown by computer, there is no human pilot required, and you could have a shift system for the “observer” operators.

Unfortunatly (for privacy) UAV’s are definatly going to be a fixture in the future, and a very lucrative market for start up companies that do not have high R&D budgets.

Clive Robinson • August 22, 2005 6:21 AM

@Ed T.

“you would have to have a fairly powerful radio transmitter to sieze control from the ground site operator”

Actually, you don’t need a great deal of transmitter power to overcome the drones very easily especially as you have no intention of controling them.

The reason for this is that due to the cost sensitive nature of the drones (and some technical issues) they will be using terestrial radio systems.

The sat systems proposed for military drones are very very expensive and with their inherant problems, are partly to blame for pilot control issues which cause drone crashes etc (think time delays of 2 or more seconds etc when trying to drive your car at >30MPH at night).

What you need as a smugler is a little observation, patience, and two fairly low power transmitters a couple of high gain antennas and a little forward planning,

1, Due to the inevitable financial cutbacks etc,

The drones that initially augmented the random ground patrols will replace them as it’s a lot lot cheaper the excuse will be “targeted patrols” / “Rapid reaction” based on intel from the drones (see stuff on CCTV in town centers).

Likewise the drone operators will all go to work in a nice office building not in moving / covert trucks etc otherwise they would want danger money or some other finacial inducement for the “front line” risk they are taking.

Also as the manufactures will want to advertise their success their broachers will have photos / write ups of the customers they have supplied (normal practice for security companies allready, again see CCTV stuff).

Then there are the operators and technicians of both the manufacture and the customer, they are very likley to be low paid civilians with similar skill sets (again see CCTV stuff). The chances are that the will talk to their buddies over a beer or three and the information will get around, along with other “war stories”.

All of the above will mean that the technical details the smuglers will need will filter into the public domain fairly quickly.

So nothing about the systems is going to remain secret or covert for very long…

All this is going to be easily obervable to teanagers (think Blue/WifI Jacking) who for the fun of it will hack the drones (they are a target just to good not to). Hey the FBI might just demonstrate it at a trade conferance or two as well…

So the tools to do it and the information required will quite safely be obtainable by the smuglers who then,

2, Identify the main control site and it’s RX antennas put a low power TX with high gain antenna pointing at it, this blinds the operators seeing what is going on. For those about to sugest normal space diversity, think again you need to do base band diversity, which has a whole heap of problems of it’s own, the solutions of which are not to difficult to design in from the start but can be very difficult to put in post production… Also it costs money not just at the design production stage but at the instalation stage as well (and customers will not pay this without good reason, remember like CCTV it’s a visable deterant not an intel gathering system).

Sadly sensible baseband diversity will probably not make it into the drone requirments analysis anyway let alone the final product specification 🙁

Even if it is designed in as a smugler you only need another TX as there is not likley to be more than two sites any way…

3, However what will make it into the requirments analysis is LOS (loss of signal) procedures.

Radio Control modlers are familier with this, you lose the signal the plane goes into a very gental dive on a circular heading, and slowly spirals to ground. It is very likley exactly the same system will be used for low cost drones, although it will probably be more like an auto pilot back to a safe “crash” site.

Public safety would require this after all you do not want a drone doing a 9-11 into an office block or school playground in a nearby town do you?

This is actually in the smuglers interests, as they will know what will happen when the drone gets LOS and will plan acordingly.

4, So based your plans on 3 you use a second low power transmitter and again a high gain antenna pointed at the drone in question by a smugerlers “operative”.

This TX transmits, random data of a similar data / frame pattern up to the drone, it overwhelms the RX front end, but more importantly the random data gets out of the RX into the data proc, where it then mucks up the software etc. Designers tend to forget that jamming margins apply at each stage down the line not just the RX front end, as this is how it’s generally portraied in text books they read when doing their degrees (if they did one).

The UK army knew this in the 1980’s when the IRA where using Radio Control equipment to remotly detonate bombs, If you send out a signal that looks just like the RC signal but does not change, either the bomb goes off when you start coming into range which is usually at a safe distance, or the bomb does not go off. Either way, your troops/assets are protected. The IRA evolved however and used a couple of small telescopes and a flash gun which was actually cheeper for them and did not have the same problems (RC equipment was effectivly a controled substance in N.I. at the time).

So in conclusion if you as a smugler do it in the right order, the drone operators first get blinded (make it intermitant in the right way over a long period of time and it will look normal to them and they will live with it). Then they lose control and the drone becomes usless.

The effected drone will be recovered, nothing will be found wrong with it so the, customer calls in the manufacture to identify the problem. Considerable periods of time will then be wasted while the manufactures technical bods look into the problem and come to no firm conclusion (other than it’s not their problem).

Mean while the smugglers now get a couple of tonnes of grade A white stuff or whatever across the boarder whenever they feel like it, all with less problems than before.

As always High tec solutions have significant problems when done on a budget, and belive me this will be done on a very small budget (see the miriad of postings about the failings of CCTV in other postings).