New Fiber Optic Signal Processing Technique Doubles Communication Distance

hypnosec writes: Researchers at University College London (UCL) have demonstrated a new technique for fiber optic signal processing that doubles the distance at which data travels error-free through transatlantic sub-marine cables. The UCL research, published in Scientific Reports, has the potential to reduce the costs of long-distance optical fiber communications as signals wouldn't need to be electronically boosted during their journey, which is important when the cables are buried underground or at the bottom of the ocean. The study reports a new way of improving the transmission distance, by undoing the interactions that occur between different optical channels as they travel side-by-side over an optical cable. By eliminating the interactions between the optical channels, researchers increase distance signals can be transmitted error-free from 3190km to 5890km, which is the largest increase ever reported for this system architecture.

Cheaper

[Bengie]

Will my Internet be cheaper now?

Re:

[andyring]

No, it just means the Nigerians can send their money requests even faster.

Re:

[Kjella]

A lot of the more remote areas of the world are islands, if you want a truly global Internet this is a good thing. Nigeria is in the middle of Africa, they can reach all of Europe and Asia without long haul ocean cables. Unless you're looking for straight lines to do HFT, but that's more of a white man's scam. This is more for crossing the Atlantic, Pacific, to Australia and so on.

Re:

[itzly]

Ocean cables are quite popular, even between places that can be connected over land.

http://www.submarinecablemap.c... [submarinecablemap.com]

Re:

[ShanghaiBill]

Nigeria is in the middle of Africa, they can reach all of Europe and Asia without long haul ocean cables.

Nigeria is connected to Europe by long haul ocean cables [wikipedia.org]. In theory, they could go across land for much of the distance, but because of political instability and lack of infrastructure, that is not feasible in practice. Dealing with Atlantic storms is easier than dealing with Boko Haram, and the Tuareg.

Re:

[NMBob]

Woohoo! No. What it means is that all of that money I keep transferring to them won't keep getting lost because of transmission errors! I'm finally going to be rich!

Re:

[account_deleted]

Comment removed based on user account deletion

Re:

[NMBob]

That's the one!! Did you get one too!? WE'RE gonna be rich. :)

Re:

[gstoddart]

No, it means they can oversubscribe now by twice as much as they do now.

NO technology will cause your internet to get cheaper -- the assholes who run telecom companies see to that.

Re:

[wonkey_monkey]

Really? When fibre optic came along here internet access became a lot cheaper per Mbps. And for those that were happy with their existing speeds, it became a lot cheaper in absolute terms.

Re:

[poetmatt]

Nah, it became a lot cheaper in some countries, notably everywhere except the US. In the US however, it was unchanged.

Re:

[Bengie]

I guess they didn't teach you how to round.

how do they do it today?

[Noah Haders]

how do they boost the signal in the middle of the atlantic????

Re:

[Anonymous Coward]

With an amplifier. It requires power over the cable. The power, for some bizarre reason, attracts sharks. Sharks chew the cable. Shark chewed cable doesn't work anymore. It's a problem.

Re:how do they do it today?

[Shatrat]

But there are still going to be amplifiers. They are referring to eliminating transponders, which we already don't use for modern sub-sea links. This is basically just 16QAM, but instead of using 4 symbols over 1 frequency, it's 2 symbols over 2 frequencies tightly spaced together. It's been understood for a while that widening the channels was probably the only way to go beyond 100gbps for a transmitter/receiver.

Re:

[hackertourist]

Do I understand the difference correctly? Transponder = optical receiver, processing equipment and laser which sends a new signal. Amplifier = EDFA?

Re:

[Shatrat]

Yes. Although more and more amplifiers are either Raman effect based or hybrid Raman/EDFA.

Re:

[Anonymous Coward]

Nothing bizarre here. Sharks (and other cartilaginous fish) have electroreceptors to hunt and track their prey, so anything generating an electromagnetic field will atract sharks to check if it's a viable food source.
http://en.wikipedia.org/wiki/Ampullae_of_Lorenzini

Re:

[Noah Haders]

so you're saying that next to the fiber optic cable there's a power cable that transmits electricity to the middle of the atlantic, where it powers an amplifier underwater? Wouldn't there be tremendous resistive losses????

Re:

[itzly]

A cable with resistive losses beats having no cable at all.

Re:

[Shatrat]

There are, which is why the electricity is at very high voltages to overcome it, which is why it attracts sharks. I'm more of a terrestrial guy, but if i remember correctly what they do is apply a very high + voltage on one end of the cable and a very high - voltage on the other end, and the amplifiers are powered in series inside of 'festoons' on the bottom of the ocean.

Re:

[Noah Haders]

what I would do is circumnambulate the globe so you don't need amplifiers. Maybe NYC to Curacao, Curacao to eastern tip of Brazil, to the ivory coast, to Gibraltar, and then to UK. Alternatively NYC to Canada to Greenland to Iceland to UK. Doesn't that make more sense?

Re:

[Shatrat]

No, it really doesn't. 5 microseconds per kilometer of latency.

Re:

[Noah Haders]

yes but how much latency do the amplifiers add? A lot, I bet.

Re:

[Shatrat]

Raman amplifiers add zero latency. EDFA amplifiers add latency equivalent to the length of the fiber coil inside, which is going to be just a few microseconds.

Re:

[Noah Haders]

mmm, raman... just 3 mins latency from the package to the pot then into my mouth...

Re:

[hackertourist]

Well, the electrical and optical conductors are in the same cable [wikipedia.org], but yes.

Re:

[Noah Haders]

what I would do is make a passive amplifier that accomplished the same thing but didn't need a power source. like an echo chamber for sound.

Re:

[hackertourist]

Passive amplifiers are perpetual motion machines. An echo chamber doesn't amplify, it merely minimizes attenuation.

Re:how do they do it today?

[msauve]

EDFA [wikipedia.org].

Re:how do they do it today?

[silas_moeckel]

Anything modern is running EDFA's where fiber is doped with Erbium and a pump laser is mixed in they get about 40db of amplification with a 100ma laser. Great part is it's not signal speed dependent as it's an all optical all analog method. The old way was to put a receiver and transmitter coupled back to back used a lot of power and was specific to a speed.

Re:

[NMBob]

How powerful are the lasers at the start of the run? mW? bunchesofW?

Re:how do they do it today?

[Shatrat]

This sort of thing is measured in dBm to make the math easier. Decibel Milliwatts, where 0 dBm = 1 mw and it's a logarithmic scale up and down from there.
Individual channels are on the order of 0 to 5 dBm, or 1 to 3 mw. The composite signal coming out of an amplifier, which consists of multiple channels, is on the order of 20 dBm depending on how many channels are active and what the reach of the amplifier is. That's about 100 mw. Definitely a laser safety concern, but not military style death ray.

Re:

[NMBob]

Got it. The 0dBm=1mw was the link I was missing. Geeze...that's not much for those kinds of distances. Pretty amazing. Thanks!

Re:

[Shatrat]

The receive sensitivity on common '10G' DWDM optics is frequently down to -24 dBm or lower. That's less than a hundredth of a mw. So from that perspective, that +0 or +3 dBm is rocket hot. 100G optics aren't quite as sensitive, but still down to -14 or -17 or so depending on the specs.

Re:

[gstoddart]

They have whales blow into the tubes ... true fact!

Discrimination

[Harold the Wombat]

"...error-free through transatlantic sub-marine cables." So those of us in Australia, which is bordered by bodies of water such as the Pacific and Indian oceans, amongst other, and NOT the Atlantic will continue to get error riddled internet. Typical Pom's, can't trust them further than you can smell them.....

Re:

[gstoddart]

So, you take a couple of empty Foster's tins, and a really long piece of string ...

Re:

[ItsJustAPseudonym]

If you take two Foster's tins that came from the same pack, aren't they quantum-entangled? I think the communications would work without the string.

Re:

[KillAllNazis]

Only if you keep your eyes closed which makes the engineering a bit more difficult.

Wait, what?

[mark-t]

... interactions that occur between different optical channels as they travel side-by-side over an optical cable...

I thought one of the major advantages of optical was that there WASN'T any interactions between adjacent fibers, unlike copper cable, where the charge of electrons running in one line can interact with the charge of electrons running in another line, and produce some interference.

Or are they talking about different communications channels on the same channel on the same *fiber*? Because tha

Re:

[mark-t]

Yes, that's what I imagined they must have meant....except such channels don't exist side by side, at all... they coexist in the same region of a single fiber.

Speed of first transatlantic cables (1858,1866)

[renergy]

Though the article is not about speed, let me quote wikipedia on that matter, as I was shocked/amazed the other day when I found how slow the communication over the cables was in the beginning:
"The reception was very bad on the 1858 cable, and it took two minutes to transmit just one character (a single letter or a single number), a rate of about 0.1 words per minute."
"..the 1866 cable, ..had been vastly improved.. could transmit eight words a minute"

Re:

[renergy]

(those were, of course, metallic cables) http://en.wikipedia.org/wiki/T... [wikipedia.org]

Re:

[dbc]

Which is when people realized there is a difference between a transmission line and just a piece of wire.

"Undoing the interactions"?

[ItsJustAPseudonym]

This phrase "undoing the interactions" is vague. It sounded like they had some kind of new principle of superposition that had been discovered. From the article:

By eliminating the interactions between the optical channels, we are able to double the distance signals can be transmitted error-free...

In reality, they are using a higher-order symbol constellation (16QAM) to carry more information per symbol, and I suspect that they combined several optical channels into a larger bandwidth to do that. I think

Re:

[lippydude]

@ItsJustAPseudonym: 'This phrase "undoing the interactions" is vague.'

The whole article is vague. They appear to be using a variant of 16QAM to create a super-channel, consisting of different frequencies channels, transmitting this and demodulating at the receiver, thereby eliminating cross-channel interference, which would occur if they transmitted each channel independently. They appear to be using a back-channel ("virtual digital journey" wha ?) to actively detect and send error-correction information