Rattle wrote:

] There are many good reasons for them not wanting anyone to
] know the details of the equipment they use for secure
] communications, but I think its unlikely that the discovery of
] the equipment would not give someone the ability to decrypt
] any communications using similar equipment.

Hrm... I'll bet they are using classified encryption systems and they are worried that if the device got into the wrong hands then enemies might be able to study it.

Decius wrote:
] Hrm... I'll bet they are using classified encryption systems
] and they are worried that if the device got into the wrong
] hands then enemies might be able to study it.

Isn't every military/government encryption system classified? Do you think they are using some flavor of PKI that the public has not their hands on? I think its certainly possible, but I also think its more likely its just RSA with a large key size or something similar, and this is the standard level of paranoia (that for good reason is) assigned to all mil/gov encryption systems.

If its a well designed system, and I'd assume it is, knowing how it works dosen't make it any less secure. The only thing that could take away from its ability to keep data transmission secret is a compromised key.

I completely understand that they wouldn't want anyone seeing it though.

Rattle wrote:

] Isn't every military/government encryption system classified?

Sort of. Fortezza isn't, but pen and paper transposition ciphers from the first world war are. Its a matter of need to know. :)

] Do you think they are using some flavor of PKI that the public
] has not their hands on?

I don't think they are using PKI. However, I KNOW they are using stuff the public doesn't have their hands on. Ask me offline if you're curious.

] If its a well designed system, and I'd assume it is, knowing
] how it works dosen't make it any less secure.

This is a fair assumption in most private sector security situations, but it is not a fair assumption in military cryptography. The more information your opponent has, the more likely it is that they will break your ciphers. We would never have cracked enigma if we hadn't gotten access to the machines themselves.

Decius wrote:
] Sort of. Fortezza isn't, but pen and paper transposition
] ciphers from the first world war are. Its a matter of need to
] know. :)

Durring Vietnam my father was stationed in Germany at a radio station where he operated some type of crypto system. He hasn't told me much about it, other then that it was an electronic system that read from ticker tape and key sequences were entered in the form of a switchboard similar to what old-school telephone operators used. (KW5? KL7? Something like that..) From what he told me, everything that had anything to do with any crypto systems was always classified, no matter how inane.

Having a ham radio license, experience fixing TVs & radios, and good timing, saved him from being shipped off to Vietnam with a rifle.

] I don't think they are using PKI. However, I KNOW they are
] using stuff the public doesn't have their hands on. Ask me
] offline if you're curious.

That being the case, getting some of the hardware could expose the cyphers..

] This is a fair assumption in most private sector security
] situations, but it is not a fair assumption in military
] cryptography. The more information your opponent has, the more
] likely it is that they will break your ciphers. We would never
] have cracked enigma if we hadn't gotten access to the machines
] themselves.

If I'm correct in my knowledge of crypto, if you understood the cipher, you would know what you would have to have your super computers do to derive a key. If you assume that the ciphers don't have any obvious weaknesses (which is safe, the guys down at the NSA know what they are doing), and the key sizes are large, then its still unlikely that it would aid you in being able to derive information from an encrypted communication within a timeframe where it would be useful.

I'm guessing that most of the unique features these non-public ciphers include have to do with key management..

It just dosen't seem logical to put something out into the field that would allow an enemy break your encryption if they found it. At least, not now that our crypto ability is to the point where a device can be smart about rotating keys, capable of keeping unique backup keys, keys being dropped if suspected compromised, etc.. In the time of enigma, those capabilities didn't exist like they do now. Same thing durring my father's time.

But then I don't know nearly as much about crypto as other people on the system. I know a lot about application of crypto technology available to me, I understand how it works, and I understand a good chunk of the math.. But thats about it..

Rattle wrote:
] If I'm correct in my knowledge of crypto, if you understood
] the cipher, you would know what you would have to have your
] super computers do to derive a key. If you assume that the
] ciphers don't have any obvious weaknesses (which is safe, the
] guys down at the NSA know what they are doing), and the key
] sizes are large, then its still unlikely that it would aid you
] in being able to derive information from an encrypted
] communication within a timeframe where it would be useful.

Your assumption is where this is breaking down. The NSA doesn't assume that the ciphers they produce don't have any obvious weaknesses. They assume that every cipher can be broken, and they assume that their enemies are at least as good at breaking ciphers as they are (which sometimes is a good assumption.)

Decius wrote:
] Your assumption is where this is breaking down. The NSA
] doesn't assume that the ciphers they produce don't have any
] obvious weaknesses. They assume that every cipher can be
] broken, and they assume that their enemies are at least as
] good at breaking ciphers as they are (which sometimes is a
] good assumption.)

I guess its a good thing they make that assumption. Better safe then sorry.