ZME Science: Yes, a quantum internet is possible, new study shows
Peter Fairbrother
peter at tsto.co.uk
Mon Dec 31 16:15:20 PST 2018
On 31/12/18 00:50, Punk wrote:
> On Sun, 30 Dec 2018 22:21:06 +0000 (UTC)
> jim bell <jdb10987 at yahoo.com> wrote:
>
>> ZME Science: Yes, a quantum internet is possible, new study shows.
>> https://www.zmescience.com/research/technology/quantum-internet-possible-26122018/
>>
>
>
> So what happens when a 'quantum' link used to send keys is
> tapped? Does the key still get through?
Yes, but.
Initially it was thought that cloning the signal (as required for
tapping it) is impossible due to the no-cloning theorem, part of quantum
physics.
Therein would lie the untappable nature of Quantum links. Physics itself
would protect you.
However it was later realised that the no-cloning theorem only says you
can't clone a quantum state perfectly - it doesn't say you can't clone
it at all.
There are several techniques which can be used for partial cloning, some
such as the well-known 1->2UQC have hard theoretical limits of 5/6
perfection - ie 5 photons in 6 can be cloned.
Others have higher limits. I think there is even one which doesn't have
any theoretical limits apart from "not absolutely perfect", but these
are harder to implement.
While it is possible to design a quantum key exchange machine to defeat
1->2UQC cloning attacks by taking advantage of the statistical
limitations of the 1->2UQC cloning process, as far as I am aware
(admittedly not very far) none of the machines in existence do so.
Plus, a secure authenticated side-channel is required to prevent MITM
attacks.
Quantum cryptography is not very useful for anything (except maybe
getting research funding).
Happy New Year to all!
Peter Fairbrother
More information about the cypherpunks
mailing list