back to article Quantum crypto - with nothing more than STANDARD broadband fibre

Boffins have worked out how to run quantum cryptography systems over a standard broadband fibre in a development that brings theoretically unbreakable encryption closer to mainstream use. Traditionally it has been necessary to use dedicated fibre to send the single photons (particles of light) that are required for Quantum Key …

COMMENTS

This topic is closed for new posts.
Silver badge

So what

It still needs to be a single continuous fibre, once it goes into a switch/repeater/router etc and is turned into electricity and back into a photon, or even directly into a new photon in a fibre amplifier, all bets are off.

8
1

Re: So what

Glad this is the first comment, it's by far the most important flaw in the model.

1
0
Gold badge
Thumb Up

So its a lock in detector with a 100ps window.

Note what's impressive is that it's detecting the presence or absence of *individual* (or very small number) of photons. Like hearing someone whispering to you from the opposite side of a football stadium in the middle of a match and *hearing* them.

And it could allow secure comms for the rest of us.

Which is pretty impressive.

4
0
JDX
Gold badge

Re: So its a lock in detector with a 100ps window.

Pretty amazing - on first read it sounds too good to be true!

0
0
Holmes

Re: So its a lock in detector with a 100ps window.

Here is the problem though... (Its only good as others have already stated) from your end point to the endpoint of the line (usually the local telco).

This solution sounds like it is great for intra-office connections, government and military uses, where they have self-run dedicated lines... not for the average Joe.

For your home, you are still having to rely that your local ISP that is bringing this cable to you has good enough security that the routers/servers that your connection is passing through is not compromised. Even further you have trust that some rogue employee is not sniffing out your communications themselves (and last time I checked most telco/ISP's don't even drug test, let alone do comprehensive background checks.

And then finally you have to trust that telco to not allow taping of the device by feds/police or just outright collecting the data and handing it over to whichever government agency requests it.

I remember when VoIP was first deployed it ran under the radar for about a year, and then all the VoIP providers (like Vonage, LIngo etc) all got a personal call from the FBI.... place these routers in your data center cages, or you are out of business....

So I don't see most Governments even allowing such lines being sold to the public (and mob, terrorists etc) without the ability of the ISP to intercept the stream of data once it is off the initial first connection.

Again, this only seems of use to me if you (your organization anyway) are in physical control of BOTH ends of the line, then you can perceive interruptions to the line in transit.... otherwise this has no value to the home user.....

Besides there are things line Tor, encryption VPN's etc that already allow secure comms for the private individual. You can protect the data stream via encryption, but you will never (via Tor or any other method) really truly hide the connection itself.... its just the nature of networks, they are trackable.

The only to not have the start and end points of your communication is to effectively disappear in the noise (not have anyone looking at you so to speak)

0
0
Anonymous Coward

Re: So its a lock in detector with a 100ps window.

"You can protect the data stream via encryption, but you will never (via Tor or any other method) really truly hide the connection itself.... its just the nature of networks, they are trackable."

A connection to a VPN provider that takes anonymity seriously and doesn't log usage is untrackable. Mind you, most providers DO log usage but a few don't.

0
0
Silver badge
Boffin

Re: So its a lock in detector with a 100ps window.

>This solution sounds like it is great for intra-office connections, government and military uses, where they have self-run dedicated lines

I doubt such connections exist outside of a single campus. Pretty much everyone relies on third parties - (ie; telcos) for long distance transport. I would expect most physical leased lines to terminate at the nearest exchange/POP and then become a virtual connection across someone else' backbone. It's just too expensive and difficult to go digging trenches or even blowing fibre across the country even for the military.

I can see verification being done at the exchange/POP at both ends though. Would that be enough? I confess I don't know enough about this stuff but if you could rely on 'the network' to be secure would be enough that the client's POP and the server's POP both agree? I guess you could verify at every switch/amp but would probably be impractical.

0
0
Anonymous Coward

Needs more work

Given that a commercial fibre pair now carries as many as 88 wavelengths at up to 100Gb/sec of traffic per wavelength - although admittedly the line rate is only about 11.5Gb/sec per wavelength - it would seem more work is needed before this actually turns into a useful technology.

0
0
Silver badge
Holmes

Re: Needs more work

"it would seem more work is needed before this actually turns into a useful technology"

Well that wins the "No Sh*t Sherlock" award for the day. You'd have scoffed at the Wright Brothers too, I'd imagine.

0
0
Thumb Down

Re: Needs more work

Isn't this to just transfer the 'key' securely though? If so, then the optical connection can be throttled back to transfer keys, and then ran again at full speed until a new key needs to be sent.

0
0
Boffin

Re: Needs more work

Erm - A 100Gbps DWDM / OTN system will send 100Gbps (plus FEC Overhead - usually between 11-16%) to line of traffic and receive 100Gbps per wavelength. No Service Provider would accept anything less.

0
0
Happy

I know what!

A Quantum Cryptography Tax should buy enough machines to decode this nonsense! CAPS inspired by AMFM

0
0
Gold badge
Unhappy

I guess it depends what the definition of a "line" is

Modern Ethenet runs into active hubs. Would this *survive* travel through such a device? If not that "point to point" is going to be pretty short.

1
2
Gold badge

Re: I guess it depends what the definition of a "line" is

As it turns out to be.

Should have read that first comment more closely.

<sign>

Back to the quantum research lab then.

1
1
Bronze badge

El Reg units please

" The Cambridge team achieved their breakthrough using a detector that is sensitive only for a very brief window (100 millionths of a micro-second)"

And what is that in double decker buses?

1
0
Alert

BOFFIN

It wouldn't be a Register article if it didn't have the word "boffin" in the title.

0
0
This topic is closed for new posts.

Forums