Fiber cut - response in seconds?

Tue Jun 2 19:41:10 UTC 2009

David Barak wrote:
> 
> Paranoia 101 teaches us that any given encryption approach will eventually fall before a brute-force onslaught of sufficient power and duration[1]. 

Of course. Hence my comment bout the likely hood of success depending on 
how much computing power they have access to. How much easier does my 
job get if I have access to thousands of encrypted e-mails vs 1 
encrypted e-mail? Once I factor your PKI root private key, your toast. 
It was my impression that the various algorithms were designed to 
prevent traffic analysis attacks, or at least vastly reduce there 
effectiveness, and if some magical corner case is discovered it should 
be further mitigated by key rotation right? I'm an operations guy, not a 
math wizard. :)

  I'm not trying to argue that the attacker in this case could 
necessarily detect a flaw in the algorithm; rather, they'll get an 
effectively infinite number of chances to bang against it with no 
consequences.  Once it's cracked, the attacker will *still* have the 
physical access which is thus compromised, and then has free access to 
all of the transmissions.

Sure. However couldn't they do this in a lab environment? Various 
botnets give them access to massive amounts of computing power on an 
ongoing basis. I presume that the folks with sufficient expertise and 
knowledge to do these attacks use exploits / back doors that ensure 
continued access to this computing power, which won't be 
detected/patched by the little tykes doing spamming/phising/data 
correlation.

Then there is the ability to buy a whole lot of specialized number 
crunching compute gear as well.

Granted the US govt has there own (classified) encryption algorithms and 
as such that can't be replicated in a lab environment and requires 
access to the physical medium carrying traffic encrypted by said 
algorithms.

> 
> Physical security is a prerequisite to all of the other approaches to communication security.  Those cases where physical security is presumed to be non-existant have to rely on a lot of out-of-band knowledge for any given method to be resistant to attack, and it's very hard to make use of a connection of that type for regular operations.

Really? The US Military uses a whole lot of wireless (satellite, ground 
baed, surface to air) links. Those links can be sniffed (by people with 
sufficient motivation/funding/gear to do so). They rely on encryption to 
protect them.