How Quantum Key Distribution Works (BB84 & E91)

[@skyvoux2686]

Very good explanation of QKD! The use of photon polarization in these protocols is called a Discrete Variable QKD, it has some downsides like requiring very good single-photon detectors, so nowadays there's a lot of research on Continuous Variable QKD (CV-QKD), which can use the very Electric Field of the signal as the quantum variable, the best part is that it's been shown to work on regular telecommunication devices, which is much more attractive for companies. I did some research in this field and it does seem like a complete CV-QKD system is almost achieved, the major problem is the speed, which is much slower than conventional communication due to the high complexity of the reconciliation protocols.

[@pizzacrusher4632]

hooray! another IM video! I’ve been eagerly awaiting it.

[@adriaancanter4573]

I don't think I could afford to pay you for that 20 part course in QM, and at 63 my mind wanders, but I'm very, very glad it wandered onto this channel, thanks!

[@EspHack]

the kind of video I have to watch again over a few weeks/months, I've missed that so thank you sir.

[@collinsokafor7208]

The best QKD video i have seen Thank you so much

[@mikedrop4421]

He has returned to drop another nugget of knowledge upon us!

[@chrisanders596]

This video is so powerful. Keep up the good work!

[@morkovija]

Highest quality content. Thank you !

[@theondono]

Wouldn’t this mean that you’re limited to the distance of a single fiber optic? If all the security comes from not being able to clone the photons, no repeaters are possible, the best you can hope for is connecting to an intermediate node and form a sort of “chain” of nodes that go from one end to the other. But now every single one of these nodes is a critical part of your security. Given that the threat to current technology (Quantum Computers) are nowhere near what they’d need to be an actual threat, this looks like a great source for academic papers with very little practical benefit.

[@ImprobableMatter]

While I agree that totally trustless communications between, say, London and San Francisco would be impossible directly, it's possible to distribute keys up to 10s of km along fibers. I think this would still be useful as an average user can communicate with a trusted party like Visa or Google. Then there is also communication with satellites.

[@theondono]

@@ImprobableMatter But that’s precisely my point, I don’t need to trust any of the intermediaries in the internet with our current technology. Future proofing our communications for fear of a technology that doesn’t exist, and has shown little progress sounds far from practical. Even *if* this solution could work without trusted intermediaries, the added cost and risk of deploying new tech would have to be weighed. Again, I’m not saying this isn’t interesting (it is!), it’s just that in every time I get into this discussion with my network engineering friends, I don’t see it happening any time soon.

[@ImprobableMatter]

@@theondono The danger is a man-in-the-middle attack if large prime numbers could be factorized (e.g. with Shor's algorithm on a quantum computer). I'm also not saying that's imminent, but certainly possible.

[@MikeRI170]

No cloning theory only forbids creating two particles with same state from one. Therefore it is theoretically possible to envision a device that will consume a photon and reemit it without gaining knowledge about it’s state (thus guaranteeing no eavesdropping) using some sort of quantum process with intermediate particles or something like this, which can be what facilitates switching and repeating for long distance connections.

[@nazaninb.khojasteh5713]

Perfectly put, thank you so much

[@forloop7713]

I have a couple of quesitons: 1.Is this type of encryption theoretically unbreakable? 2.When the satellite recorded the signal sent to it by China, shouldn't that signal be stored in the satellite's memory, a memory that can be read. Thanks!

[@qwoop6351]

1. More than encryption this is communication, a way to handle key exchanges without evasdropers or Man in the middle stuff. However it enables use of One time pad remotely and safely, which would in effect be stronger than current public/private key stuff due to lacking the (very elaborate) mathematical tells. 2. Yeah but as explained in the example on the video with the perfect photon replicator, even if someone is able to read the memory/clone the messages you would still be able to deal with it

[@ImprobableMatter]

Yes, it is unbreakable theoretically. Yes, you are establishing a link to the satellite, so if someone could physically intercept the orbiting satellite they might be able to hack it somehow.

[@katinkajako4475]

This is a very good video!

[@jolietjake19]

In the E91 method, how do the photons get entangled in the first place?

[@ImprobableMatter]

One way is to take a laser, which produces photons and then use a crystal which converts each of them into two other photons with half the energy. In that process, each pair produced are entangled. In reference [7] in the description, they use a β-barium-borate (BBO) crystal.

[@chengong388]

Wouldn’t the atmosphere or molecules in a fibre optic cable mess up the superposition?

[@CrackheadMagnate]

Hey @imporable matter, will u do a sparc video in the future ? I saw it had a Q of 11 but I'm not a fusion scientist so I came here

[@leilanicatrose3576]

Sorry but this is quite confusing at 8:15 , why on earth will bob receive a diagonal to right photon, if the one sent by the eavesdropper is vertical, surely he should receive a vertical photon sent by the eavesdropper?

[@ImprobableMatter]

Yes, he *receives it* in the vertical, but since he is measuring in the diagonal basis, the only measurement options are / or \.

[@leilanicatrose3576]

⁠@@ImprobableMatter oh okay, so bob can still measure any polarised photon, but gets the reading simply based on the filter that he has used at that time. Thank you for clarifying!!

[@theonlyhero]

If a quantum internet is required doesn't that mean, that one would have to have one dedicated fiber optics connection between every single receiver? so for a few actors, that would be possible, but even just for every country a connection to every other country would be [assuming 200 countries and if im not wrong this should be combination statistics = 200!/(2!*(200-2)!) ] 19900 connections. this seems impractical. Or am i missing something?

[@ImprobableMatter]

If you want a super-duper secure one, yes. But you could probably trust Google etc, so people would only need to connect to a few fairly local servers belonging to trusted parties. I mean, you have to trust whoever manufactured your CPU, your RAM and so on anyway.

[@theonlyhero]

@@ImprobableMatter Google has currently plans for about 50 data centers worldwide. If we assume 3 other companies have the same (e.g. microsoft, Facebook and apple or so) then the number isn't lower. Or am I missing something?

[@ImprobableMatter]

That's fair, but the point is that you don't need a dedicated fiber optic between every node/datacenter. They could all share a common quantum fiber network (just like they do now) and they would detect if someone was trying to eavesdrop. For long distances, satellite constellations could be used. A single polarization analyzer is also quite small, so it would be reasonably cheap to distribute many "Bobs" around the world as well.

[@theonlyhero]

@@ImprobableMatter ah ok, thx!

[@Danji_Coppersmoke]

BB84 (7:00) basis exchange ONLY AFTER information bit HAD BEEN sent... even if both knew their basis ahead of transmission. ( just my 2c.. no expert.. I first confused with thought what if middle man got all those basis and measure it... but then there is nothing to measure since all photons had been sent... maybe I am just slow learner.. lol) .. great explanation... I knew nothing before .. now a little.. lol

[@kamamuradkhanli7854]

thank you so much for this detail it got me confused

[@andrewnelson4148]

Could we get a future video about public key encryption?

[@ImprobableMatter]

It would probably be a livestream if at all. I'll get a blackboard and go through RSA as an example.

[@andrewnelson4148]

@@ImprobableMatter I don't understand how we would do asymmetric encryption with one and zero. Will it be explained at that level?

[@TerryMack-co7lt]

Couldn't a Eavesdropper run an attack where they basically DoS attack connections just by observing them? EDIT: You mention this in your video actually. How hard would it be to run such an attack and how could you be caught? If its too easy to deny connections by observing it would render the method useless I would think.

[@ImprobableMatter]

A single eavesdropper node in the quantum network absolutely could do so - just like a malicious node could mess with packets on the ordinary internet. The point is to have a network with enough faithful nodes so that you could find a working route - again, like the classical internet.

[@ThePerceivEdLearningHub]

What is the software you are using for creating this content like the animation and white-broad?

[@ImprobableMatter]

The vast majority of graphics are made using Inkscape. I export multiple layers as images with each of the objects, like photons in this case, and then move them around in the animation software. Any free or commercial video editing software can move, fade, rotate and so on if you have images.

[@ThePerceivEdLearningHub]

Thank you!!

[@Atlas_Enderium]

What’s the point? You didn’t mention how lattice-based cryptosystems are post-quantum resistant and don’t require the facilitation of communicating with qubits, just normal electronic communications. They don’t depend on the quantum-prone prime factorization or discrete logarithm problems but rather on the mathematical hardness of the shortest vector problem. While these lattice-based cryptosystems haven’t been fully adopted or standardized yet, the underlying mathematical hardness is proven to be resistant in a post-quantum landscape, completely side stepping the need for new communication schemes involving quantum qubits.

[@Pentagram666mar]

One thing i don't understand is to how change our internet etc into photon based

[@ImprobableMatter]

I had mine changed to photon based recently. We had a fiber optic connection put in.

[@Pentagram666mar]

@@ImprobableMatter the answer is way easier than i imagined :P

[@s-tierbeers7778]

New IP video and I drop everything!!

[@Dogo.R]

Detecting an evesdropper feels a bit... secondary. I want to know no one can evesdrop. Not whether someone did. I want to aim to send a 100% secret message and have both parties know for sure its from each of us. But without needing to reprove identity frequently.

[@monsterthenergydrink]

The point is that if you can detect an eaves dropper then you send a test message before the real one which is why the DDoS attack still works but you won't ever have the content of the message or information you want to send actually stolen, so detection is pretty much just as good as just sending the secret message

[@TherealElijhawatkins]

11:15

[@BubblesTheAlphaWhale]

This might seem a silly idea for a video, but I was too young to understand the craze that happened surrounding the concept of Cold Fusion. Could you explain what they hell it was, and how it would have supposedly worked. I don't personally believe such a thing could exist , but trying to read into it and understand why it received so much backlash is a nightmare to unpack that I just can't make any sense of.

[@4thjulybd809]

So when they got them wrong but nobody was listening, wuold they think someone was listening?

[@ianthehunter3532]

Looks like someone is a fan of cartoon shows 😄1:58

[@user-ki4bg4ke9g]

Hell yhea new video

[@charlesdrury9712]

Fiber optic would not affect Because if it did the Internet we shut down

[@4thjulybd809]

How the hell do you have 194 visualisations, 2 comments and 34k+ likes???? Edit: in all your videos you have exactly 34.060 likes?? What is going on here?

[@CarolineFeliciano-k3f]

Abbott Mill

[@SomaJustinio-v2m]

Kade Lights

[@4thjulybd809]

Just noticed you have 34.060 likes in the latest video. Is this some youtube bs?

[@vik24oct1991]

Method is very flimsy for practical use, a eavesdropper can make the network unusable by just reading the bits unlike in classical networks where eavesdropper can read data but it will remain untouched, so the communication can work even if there is passive snooping going on which is the most way of eavesdropping, yes active snopping will have same effect as passive snooping in quantum secured network but its very rare as you need more privileges to do it.

[@ImprobableMatter]

They could make a classical network unusable too, just by cutting the connection they are accessing. The point is to have a multiply-connected network so that a few malicious nodes do not crash the whole system.

[@bekaltun]

.

[@MollyWi]

Fttttp... Polarisation again… . Polarisation is not superposition… It is either polarised in a direction or not directional polarised. When you watch a 3D movie, is it is a superposition? The eves dropper can simply re-apply the same polarisation and nothing is disturbed (not that it would have been anyway). Normally with these so called 'tests', they fire 10s of thousands of photons with a large degree of noise, and then hypothesize that in a noise free environment it should be possible.. Sharing the bases, does not make any difference either… Was this written by someone who has never worked with encryption?