backtop


Print 13 comment(s) - last by drycrust3.. on May 29 at 12:42 AM

By mirroring beam in line, noise is cancelled

It is an elegant and simple solution, but one that is relatively new to the world of fiber optic signaling -- beam mirroring.  The approach, already frequently used in the noise-cancelling headphones, is being touted as a promising new route to improving fiber optic routing in a new study by researchers at telecommunications company Alcatel-Lucent SA's (EPA:ALU) Bell Labs unit.

The idea involves making "twins" -- dual beams of light in a fiber that mirror each other.  Each peak in one beam is a trough in its twin.  Together they bounce along the line, much as a single beam would.

One major limiting factor in fiber optics is noise.  In order to send signals farther, beams are transmitted at higher power.  But the higher the power, the more beams tend to interact with the material in the fiber's walls, adding noise.  Beams have a limiting maximum distance they can travel and maintain fidelity -- after that they need to be received and rerouted, hopping along the next link.

By adopting the "twin" approach, light can travel four times farther that a single beam could.  In their study, the Bell Labs stream piped data at 400 Gigabits per second (Gbps) -- four times faster than the best commercially available speeds, down 12,800 kilometers (7953 miles) of fiber.  That's a longer line than the longest transoceanic link.

While note the first study to suggest the phase conjugate approach, Bell Labs claims its work offers the most straightforward implementation and is proven to travel farther without rerouting.  That means the need to reroute or boost signals during long transoceanic links may no longer be needed.

Fiber optic cable
[Image Source: Guardian UK]

Lead author Xiang Liu of Bell Laboratories in New Jersey comments in an interview with BBC News:

Sometimes you may send data from London to New York, sometimes you may send it from London to Paris. The links are changing and you cannot keep sending people to the middle of the link.

At the receiver, if you superimpose the two waves, then all the distortions will magically cancel each other out, so you obtain the original signal back.  This concept, looking back, is quite easy to understand, but surprisingly, nobody did this before.

Nowadays everybody is consuming more and more bandwidth - demanding more and more communication.  We need to solve some of the fundamental problems to sustain the capacity growth.

The approach may allow faster data transfer speeds too.  As it reduces signal noise, it allows for less repetition of information in a given beam.

The study on the work was published in the prestigious peer-reviewed journal Nature Photonics.

Alcatel-Lucent in currently in the midst of constructing a 100 Gbps undersea fiber link between Malaysia, India, and the Middle East.

Sources: Nature Photonics, BBC News



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

2 separate frequencies?
By sixteenornumber on 5/28/2013 3:04:06 AM , Rating: 2
Can someone explain this further? When they say they are sending two beams mirroring each other, do they mean two beams with different center frequencies?




RE: 2 separate frequencies?
By integr8d on 5/28/2013 4:36:12 AM , Rating: 3
It sounds a lot like the way XLR audio cables work. It's electrons vs. photons. But the concept with the audio path is that one signal is the straight signal, while the second signal is the inverse. Since they're both traveling along the same line, they'll pick up the same noise. As they reach their destination, the inverse signal is flipped (to mirror the straight signal) but the noise in the inverse signal is also flipped 180 degrees out of phase. The noise cancels itself out, while only the audio signal goes through cleanly.

Source:

In phase=====> audio * noise * audio * noise...
Out of phase==> (-)audio * noise * (-)audio * noise...

Destination:

In phase=====> audio * noise * audio * noise...
Flipped======> audio * (-)noise * audio * (-)noise...


RE: 2 separate frequencies?
By Solandri on 5/28/2013 5:43:06 AM , Rating: 1
Integr8d's gave a nice written description, but sometimes a picture is worth a thousand words. Apple has a very nice description with pictures of how it works.
http://documentation.apple.com/en/soundtrackpro/us...

XLR cables and balanced phono cables use this trick, but its primary use is in UTP - unshielded twisted pair. Pretty much every phone and ethernet cable you've ever run across uses this trick to reject noise. That's the reason they're twisted - to ensure the pair of wires stay right next to each other no matter how you coil the cable, so they pick up the exact same noise.

Audio cables are usually shielded, but this trick works so well that if you're careful to match up the right signal pairs, you can solder an XLR cable to UTP cable (which is not shielded) and it'll work just fine. I've run balanced audio through >100 feet of cat5 network cable with no audible degradation. I've heard of people doing it to 500 feet without problems. The vast majority of the noise rejection is from the balanced signals, not the shielding.


RE: 2 separate frequencies?
By PaFromFL on 5/28/2013 8:22:37 AM , Rating: 2
I'd guess that the "noise" actually refers to the interaction between the fiber and the photons, which depends on the photon electric field strength. They might be referring to power losses or nonlinearly-induced "harmonics". The two signals are probably designed to "cancel" so that the maximum electric field is reduced within the fiber medium.

With electrons (fermions), noise can be induced by external electric and magnetic fields. Electrical cables can be designed to nearly cancel external fields. Photons (bosons) have the advantage that they are almost unaffected by external fields. However, there is a limit on the photon power level because very strong photon electric fields can interact with the fiber molecules.


RE: 2 separate frequencies?
By PaFromFL on 5/28/2013 1:57:35 PM , Rating: 2
After looking at the BBC link, which also doesn't reveal any details, it sounds like they might be sending the out-of-phase signals over two similar fibers. By subtracting the received signals, they double the signal amplitude and quadruple the signal power. This would cancel "noise" components associated with even powers of the nonlinear response, but not odd powers. The "noise" they are talking about is probably signal components derived from nonlinear distortion that are correlated with the signal. More often, noise refers to random interference that is not correlated with the signal.


“And I don't know why [Apple is] acting like it’s superior. I don't even get it. What are they trying to say?” -- Bill Gates on the Mac ads














botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki