Major Malfunction makes the news again

[renderman]

Re: Major Malfunction makes the news again

What's the old adage; "Never attribute to malice what you can attribute to stupidity"?

We know that security never perfect. The goal should not be 'un-breakable' but layers that compliment each other and frustrate an attacker for a suitably long period.

In the case of crypto, having an algorithm that can't be brute forced for 20+ years when the useful lifespan of the data is 10 years is an example. A simpler one is having a safe that takes 30 minutes to manipulate, but having a guard come by every 10 minutes.

In the case of contact based passports, they have a fixed valid lifespan of 10 years, that's the standard to build to. I'm not saying that it's a flawless solution. Power analysis is a facinating field, and Adi Shamir has been doing some great work with RFID on that. However, direct contact would still be needed to read power levels to get the keys for a single passport. If the data is externally encrypted before being written and no data keys are stored on the passport, an attacker could possibly extract the data, but be up against a second problem of decrypting it.

Given gov't history of not properly securing the keys to the kingdom, it is likely that the reader keys could be compromised, however, this still only helps the attacker if they get physical possession of the passport. As well, if you have the ability to revoke keys (i.e. update your passport on the way out of the country, at passport office, post office, whatever), you can maintain the system through the 10 years or more and not have to re-tool each time someone screws up. How much do you think it's going to cost in 2 years for the next epassport version once this one is scrapped for whatever reason?

The contact based solution basically prevents the wholesale collection of passport information and mitigates alot of the passive attacks. It adds the additional layer of a harder to forge, crypto signed 'page' that has to match up with everything else, but reduces threats to classic problems of pick pocketing, mugging, loss, etc. Your left with classic, well understood problems rather than new unforseen ones.

[VAX_to_PBX]

Re: Major Malfunction makes the news again

I'd save my self some trouble, and go with Transponders. They have no BUS traces to analyze, and no hex dump functions.

Texas Instruments probably have the best line of glass transponders. Transponders also have a really good track record in the automotive industry. The only vulnerabilitys where results of other components. The only successful direct attack ever where by some college students who basically setup a GATE array to do the real time calculations.

Even if you could do a man in the middle on it you'd have to have a lot of dedicated computing power on the spot with the addition of a short wavelength radio. You'd still get impurity's though because the signal is low yield.

It's also a lot smaller than any chip with the same functionality. It'll fit in a locker key head.

[Voltage Spike]

Re: Major Malfunction makes the news again

I'd save my self some trouble, and go with Transponders.

From what I've read, they're RFID ... but they're more obscure, don't deal with collisions well, and, if it matters, are powered by a slightly different energy source. So they're better than "second-generation" two-way authenticated RFIDs how?

I still agree with others in this thread: wireless is a silly pursuit for this application. If the authorities must have physical access to the passport in order to verify, why is it such a burden for them to hold the passport against a physical-contact system?

Then there's the fact that authorities can't trust any of the data they're pulling off the passports in the first place...

[VAX_to_PBX]

Re: Major Malfunction makes the news again

I agree about Wireless, but transponders aren't like RFID that stores formatted data. They respond with bit streams with the functionality of polymorphism. As for the questioning there security. The proof is in the metaphorical pudding in this case. Just look at the results of security research on them by any 3rd party you care to mention. Specifically in vehicle's where they respond to checksums by the cars control modules.

Even in the best attack senerio known they put up a good defense against dedicated computing.

WIFI goes along with Biafras' Law "Give me convienance, or give me death." Of course you probably live in suburbia, so no need to explain. Needless to say a system that requires patience wouldn't be practical in todays world.