Slashdot Mirror
Temporal Cloak Erases Data From History
ananyo writes "Electrical engineers have used lasers to create a cloak that can hide communications in a 'time hole', so that it seems as if they were never sent. The method is the first that can cloak data streams sent at the rapid rates typically seen in telecommunications systems. It opens the door to ultra-secure transmission schemes, and may also provide a way to better shield information from noise corruption (abstract). The researchers manipulated laser light in time to create regular periods with zero light intensity (a Talbot carpet) in which to hide data. Unfortunately, the current set up erases the data-adding event entirely from history. Though they are confident that future modifications will allow them, or others, to send secret messages successfully, the more immediate use of the technology will be to cut down crosstalk when multiple data streams share the same fibre." Also at Slash Datacenter.
Yep, all that mumbo jumbo about time cloaks comes down to this:
They found a way to turn a laser on and off really fast, and at the other end of the fiber undo it so it appears to have stayed on. The whole "cloak" thing is just the idea that while the laser is off, some other signal could be sent on the fiber. Yay, with more refinement they can use it to send two channels on one fiber. The current implementation isn't able to read the second channel.
In theory, you could cut into a backbone provider's fiber and insert one of these transmitters that adds a second channel. At the other end of the fiber, you could insert the "undo" unit, so the owner of the fiber couldn't see your signal. You'd be using their fiber without their knowledge. Of course, a fiber equipped router at each end would achieve the same result.
I never did much time in college either but, that's besides the point. I found a couple of better articles and exploitations, by people who seem to actually understand it:
http://www.technologyreview.com/view/424682/first-demonstration-of-time-cloaking/
So, they are speeding up some photons, and slowing others, to create a "gap", passing something through that gap, and then, readjusting speeds. So, imagine the beam is..... Route 1 in Saugus MA. One of my favorite roads. Its not just 6 lanes of bumper to bumper traffic, that traffic is bumper to bumper at full speed.
Like most beams of light, you have fuck all chance of passing through it without casting a shadow (this is the detected event being "hidden"). But imagine if all the cars were in communication by computer. A mile down the road about haldf the cars speed up, and bunch together, and the other half all slow down, then all resume normal speed, creating a traveling gap.
Now, if you knew this gap was coming, you could scurry through it without traffic detecting you across the pavement.... after which, they perform the opposite operation, sealing the gap, as if nothing happened.
Sounds like a really cool system doesn't it? Now lets imagine it has a limitation of a 100 foot gap, moving at 75 MPH, giving you less than a second to pass the 60 feet of tar before you get....detected
I think part of the reason this sounds so weird is the terminology. It makes sense, there is no point in space where an object can be continuously and not be detected by the beam, however there are points which the beam intersects in space where the object can be, between the beam, and not interfere with the final beam...
Or at least, that's what I get from it.
"I opened my eyes, and everything went dark again"
Because the light pulses went into the pipe. Theoretically, if the build their modulators right and exactly undo what they did to make their "time holes" (which normal people would call dark intervals the original unmodulated light will be there on the fiber, superposed with a temporally spread version of the coded pulses they injected, which would be there, but hard to detect.
At the other end of the obscured connection, the trick is to run the same modulation scheme again exactly and if you do it just right, the monochromatic component of the light will once again be spread into the gappy modulation pattern you had at the first modulator output, and the pulses you injected will be there in the otherwise dark intervals.
The scheme depends on two things: (1) you have to time it exactly so that you don't lose the phasing between the original modulator pair and the modulator at the receiving end. If you miss the phase, your data will be in the bright intervals instead of the dark intervals and you won't be able to read it easily. That's what apparently happened in this experiment and why they're confident that what they did really happened even though they didn't get their data back out of the system. The other thing (2) it depends on is that the modulators have to be linear enough that when you modulate and remodulate the light, you don't mix the spectra of the data with the stream you're using to obscure the pulse sequence. If they mix, you won't get a clean signal out no matter how exactly you match the modulators.
The second is a likely limitation on this scheme because they are using nonlinear modulators -- all electro-optically active materials are nonlinear. Linear materials don't modulate light. To do completely linear modulation, you would need to do the modulation mechanically, which is much to slow to create the kind of quite intervals they need to obscure their data.