Quantum Cryptography , first is a technique to do encryption using quantum properties of elements . ( something like Pauli's exclusion theory ...we can not have two spin momentum of same nature ...nothing to worry much on this) since it is going to do at quantum level we expect the speed of encryption very fast as well as its strength. That is to say it encrypt huge data in less time .( A common problem when we use Mathematical process). But why it has less data rate and something like distance 90 kms involved. the question is here the data is represented in quantum charecteristics , hence to maintain the same for a longer distance may not be tenable also the key changing data rate can not be large due to this.

hope it helps.

That's an excellent question. We are talking about quantum key distribution here; the usual situation that researchers consider is that the quantum communication is *only* used to distribute key material and then the key material is used in a conventional encryption system. So the data rate needed for the key is much less than the data needed for the actual encrypted data. How much less? I guess that depends on how large your key is and how often you want to change it. In the most secure system, the one-time pad, your key is as long as your message so you need key rates to be as fast as data rates. More standard grade systems, like AES, require an almost trivial amount of key material, like 256 bits every month, in which case the key rate hardly even matters. So the answer is,, as usual, it depends. :) Hope that helps!

Thank you for your answers!

It seems to me, as someone who isn't an expert, that there is an important distinction between quantum cryptography and quantum key exchange. As I understand it, Mr. Prasad is talking about the former, but this paper is about the latter (as Ph.D. Holden describes). I think it is quite clear that an OTP is not feasible, because the bandwidth the authors treat is on the order of gigabits. Given that we can use this key exchange mechanism for a generic symmetric cipher, the next question that arises would be how to practically apply it. As I understand it, the key distribution scheme presented here is aimed at providing key exchange over a single fiber. Can we establish a secure connection when there is an intermediate node? For example:

Alice switch (medium access layer) Bob

Alice router (network layer) Bob

As I understand* quantum physics, it is not possible to 'forward' particles in the way that we would forward packets. This would thus require a switched network (that is, like the telephone used to be), which would severely limit the applicability of this work.

* as noted, I barely understand quantum physics. I base this on the principle that one cannot duplicate a particle's state. Please correct me if I'm wrong.

Thanks again!

Hi Rens,

You've hit on the exactly key principle behind QKD, which is that one cannot perfectly clone quantum states. This is what makes the whole process work (i.e. it is this limitation the defeats the would-be eavesdropper) but this also means traditional forwarding is impossible. Thus current QKD schemes are limited to single hop transmissions which at the moment is only feasible over ~100km.

To go further distances one would have to either use a free space uplink to a satellite, or build genuinely quantum nodes. In the literature these are called quantum repeaters, but at the moment they exist only in theory and some proof of principle experiments for some of the necessary components.

Also I should emphasise that if one wants complete security OTP is the only option. In the papers you quote the authors aren't talking about sending Gbps of key, rather they are showing you can send ~66kbps through a fibre that also has Gbps of classical communication traffic.

Hope that was useful,

Nathan