Slashdot Mirror
Single Photons Bounced Off Orbiting Satellite
KentuckyFC writes "If we're ever going to benefit from the perfect security of quantum communication, we're going to need ways of transmitting entangled photons around the globe and certainly further than the current record of 144km through the atmosphere. Anton Zeilinger at the University of Vienna and colleagues have taken an important step towards this by bouncing individual photons off the Ajisai geodetic satellite (essentially a space-based disco ball) which is orbiting at 1400km. The group says the experiment is an important proof of principle for satellite-based quantum communications."
Intercepting (and breaking) quantum crypto is very much so possible. If Charlie intercepts Bob's fotons on their way to Alice, and Charlie can transmit the very same fotons he just recieved, he can intercept the message succesfully without Alice or Bob ever noticing (perhaps a lag because Charlie has to do some work before he transmits). What Charlie cannot do is old style wiretapping: every foton is a carier of one bit and reading it causes the bit to flip and thus Alice knowing the line is tapped. Reading the foton, throwing that foton away and sending another foton with precisely the same orientation/spin as the foton he recieved to Alice will be undetectably. Never mind that in practice not now and probably not you can have 1 foton = 1 bit, but you'll need a group of fotons = 1 bit. And when there's a group, there's room to play for a wiretapper.
I refer you to the no cloning theorem. And you are correct, that in practice there is a probability that there can be more than one photon per bit sent. However, there are ways of correcting for this such as privacy amplification. The guys who are doing this stuff are smart. They know what they're doing. Physicists don't throw around terms like "provably secure" lightly.
The German Enigma machine from WWII was fairly uncrackable, even for the decoders at Betchley Park, but the German operators got lazy, and weren't following proper procedures, which allowed Turning et. al. to get a toehold into the crack.
The One Time Pad is provably unbreakable, but the British were able to decypher Soviet OTPs, because they had reused the pads after a year, thinking no-one would go back that far. One of our Admirals did the same thing, but there's no evidence he got caught.
The obvious back door to a quantum crypto machine would be the operator, or some part of the process before the plaintext is encrypted. Just because physics says that the process is unbreakable, doesn't mean the machine can't malfunction, and the malfunction may not be detectable by the operator. i.e. perhaps the machine can detect any malfunction, but the indicator light that tells the operator is broken.
Even though OTP is provably secure, the military proceeds and ends each message with gibberish, to throw off attempts at decoding.
All ideas^H^H^H^H^Hprocesses in this post are Patent Pending. (as well as the process of patenting all postings)
Oh great...
I can't wait until someone at the RIAA figures out how to protect music with quantum DRM. You get to listen to a song ONCE, then it doesn't exist anymore.
They will charge PER listening.
- For the complete works of Shakespeare: cat
Others have created quantum crypto systems that take the possibility of cloning into account, http://w3.antd.nist.gov/pubs/Mink-SPIE-One-Time-Pad-6244_22.pdf
'basic' quantum cryptography that is taught can be hacked This is true but I think not for the reasons you believe. Basic quantum crypto provides confidentiality only. To keep from being hacked, you must provide authentication as well (Alice must be able to prove she is communicating with Bob and not Eve). I haven't heard of a way to do this without falling back onto more conventional cryptographic techniques such as RSA signatures - at least when doing quantum crypto over fiber. Maybe sending photons through the atmosphere means you can actually just see if somebody is acting as a man-in-the-middle.