Untappable quantum communications?

[GTOM]

Could please someone explain me, if entanglement can't be used to direct communication (that would be an FTL dream come through) how untappable quantum communication supposed to work?
If regular data transmit still needed to get any useful information out of the entangled particles, why it can't be tapped? What if someone could get acces to entanglement / distributer / repeater node itself?

 

[phinds]

Could please someone explain me, if entanglement can't be used to direct communication (that would be an FTL dream come through) how untappable quantum communication supposed to work?

It isn't, if you are talking about using entanglement for communication. There IS no meaningful / useful communication via entanglement

If regular data transmit still needed to get any useful information out of the entangled particles, why it can't be tapped?

What do you mean? It CAN be tapped, if you are talking about the non-FTL communication and it doesn't exist if you are talking about FTL communications.

What if someone could get acces to entanglement / distributer / repeater node itself?

As far as I can tell, that sentence makes no sense.

 

[Nugatory]

If regular data transmit still needed to get any useful information out of the entangled particles, why it can't be tapped?

Alice and Bob use entanglement to create an encryption key that is known only to them. Once they have that key, they use ordinary symmetric-key encryption to send their message through the regular channel. Eve, the eavesdropper, can tap the regular channel, but without the key she cannot get any information from it.

What if someone could get acces to entanglement / distributer / repeater node itself?

They would be able to shut the key-generation system down, but that's all. They would not be able to manipulate it in such a way that they could learn the keys that Alice and Bob are generating.
(Quantum key distribution is vulnerable to a man-in-the-middle attack, but this attack requires control of the classical channel).

 

[DrClaude]

The point of quantum cryptography is not to make eavesdropping impossible, but fully detectable.

 

[GTOM]

Alice and Bob use entanglement to create an encryption key that is known only to them. Once they have that key, they use ordinary symmetric-key encryption to send their message through the regular channel. Eve, the eavesdropper, can tap the regular channel, but without the key she cannot get any information from it.

They would be able to shut the key-generation system down, but that's all. They would not be able to manipulate it in such a way that they could learn the keys that Alice and Bob are generating.

Thank you. But i still don't understand it fully.

So basic 1 : quantum information can't be cloned, any measurement change it, and we can't get all information in a single measurement
basic 2 : in case of entanglement tampering with one side means the other side instantly (or at least faster than light, unknown max distance) reacts, but next point
basic 3 : no matter how great it sounds, we CANT use it for signaling, since coherence is destroyed instantly, and before measurement we can't be even sure whether it in is state A or B, so it means nothing if we find 2 particles in state A and 3 in state B

I hope i managed to get the basics right.

So how eavesdropping is detected, how do we know, that a spy tampered with the particles, and not our measurement broke the coherence? The information is changed, but doesn't entanglement means it changed on both side?
Is it because info teleportation, that we add info C to an entangled pair, and it appears on the other side if coherence still stands? And even if the other side don't know, that whether the result is A+C or B+C or A+D (still no signaling useful data) we can count on that both sides have the same key?

What about noisy quantum key distribution channel? That could destroy coherence too, so redundancy is needed, is there a clear difference between tapping and simple noise?

(Quantum key distribution is vulnerable to a man-in-the-middle attack, but this attack requires control of the classical channel)

What does exactly vulnerable means in this case, and what the attacker could achieve?

 

[Nugatory]

You'll find much good stuff if you google for "quantum key distribution"; the wikipedia article is a good start. But some quick answers:
1) If there is any tampering Alice and Bob will end up creating different keys. This becomes apparent the first time they exchange messages, but no sooner. There is no way of detecting tampering or attempted eavesdropping without transmitting information on the classical channel.
2) The man-in-the-middle attack is the same as for non-quantum cryptography. If Mallory controls all communication channels between Alice and Bob, he can impersonate Alice and he go through the key exchange protocol with Bob, and vice versa with Alice. Now when Bob believes that he's sending a secure message to Alice encrypted with the key he negotiated with Alice, he's actually sending it to Mallory, who decrypts it, reads it, then reencrypts it using the key he has negotiated with Alice and sends it on to her.

 

[GTOM]

You'll find much good stuff if you google for "quantum key distribution"; the wikipedia article is a good start. But some quick answers:
1) If there is any tampering Alice and Bob will end up creating different keys. This becomes apparent the first time they exchange messages, but no sooner. There is no way of detecting tampering or attempted eavesdropping without transmitting information on the classical channel.
2) The man-in-the-middle attack is the same as for non-quantum cryptography. If Mallory controls all communication channels between Alice and Bob, he can impersonate Alice and he go through the key exchange protocol with Bob, and vice versa with Alice. Now when Bob believes that he's sending a secure message to Alice encrypted with the key he negotiated with Alice, he's actually sending it to Mallory, who decrypts it, reads it, then reencrypts it using the key he has negotiated with Alice and sends it on to her.

All communication channels between Alice and Bob, AND the quantum key distributor unit, do i get it right?Otherwise i found this one.

https://phys.org/news/2016-10-precise-quantum-cloning-pathway.html

Maybe quantum information cloning isn't that impossible? (Considering, that one has to calculate with regular transmission errors.)

 

[Misha]

All communication channels between Alice and Bob, AND the quantum key distributor unit, do i get it right?Otherwise i found this one.

https://phys.org/news/2016-10-precise-quantum-cloning-pathway.html

Maybe quantum information cloning isn't that impossible? (Considering, that one has to calculate with regular transmission errors.)

The problem of cloning is traditionally badly covered in discussions of QKD. Usually, the no-cloning theorem is shown for a general case, which is almost trivial, while the case of finite sets, the most relevant and interesting but difficult, is silently omitted.

There's no problem in distinguishing orthogonal states and, therefore, states from such finite sets can be easily cloned. Once states are not orthogonal, however, the best one can do is to clone a state with some probability and to know when cloning failed. If in QKD failed attempts are substituted by random states, this will show up as noise in subsequent key distillation. Thus, noise can be tolerated only up to a certain level, above which the communication should be regarded as insecure. This is the standard approach.

Nevertheless, no-cloning is still there and not going anywhere. It's a shame that the trend in public releases, and sometimes original publications, is to hide this fact.