Re:Pentagon Paying China To Carry Data
by quanticle on Tuesday April 30, 2013 @01:36AM
(#5630581)
Isn't this exactly the sort of problem public-key cryptography was designed to solve?
[ Parent
| on HN ]
Signals intelligence is not just about reading enemy communications. Traffic analysis[0] would give the Chinese valuable insight into US military communication patterns. Changes in data patterns could alert them to changes on the ground.
What stops the US from sending a whole bunch of constant dummy traffic to disguise this? I realize bandwidth capacity and cost could be a reason, but if there were enough capacity and budget? OK, maybe that question is a non-starter.
I'm reminded of a funny time when I was in an internship. My team and another team was fighting over disk space in the database servers. So we created a whole bunch of dummy tables. Whenever one of us needed to do something, we'd request for one of the dummy tables to be dropped so that we could do our thing. The other team could never do anything. It was weird.
Re:Pentagon Paying China To Carry Data
by michael_miller on Tuesday April 30, 2013 @01:53AM
(#5630612)
Cryptography solves the problem of spying on or altering the communications, but not the problem of blocking them. If the US was engaged in a conflict that China didn't like, they could theoretically cut off satellite communications, rendering our drones useless.
Re:Pentagon Paying China To Carry Data
by wereHamster on Tuesday April 30, 2013 @01:50AM
(#5630604)
Or cryptography in general. Doesn't need to be public-key, could just as well be symmetric crypto.
Well, sure. The one big advantage that public key crypto brings is that you don't need a separate channel for key exchanges.