[LINK] 400G fibre connections

[stephen at melbpc.org.au]

. Scientists build 400G connection

By Jon Gold (Network World)29 May, 2013
<www.arnnet.com.au/article/463161/look_google_fiber_scientists_build_400g_c
onnection/>


A research team led by Bell Labs' Xiang Liu has published an article in 
Nature Photonics describing a way to send and receive information at 
400Gbps across 12,800km of optical fiber, an enormous potential gain of 
both speed and effective distance compared to current technology.

<www.nature.com/nphoton/journal/vaop/ncurrent/full/nphoton.2013.109.html>

The idea was likened, by the BBC, to the wave-canceling technology used in 
headphones that block outside noise: Ambient soundwaves are detected by the 
headphones, which then play back an inverse waveform to cancel them out.

However, Liu tells Network World that the fiber-optic technique his team 
used is at least in principle actually simpler than noise-canceling 
headphones.

"In noise-cancelling headphones, one needs to detect the noise and then 
remove it from the signal, while in the twin-wave case, we do not need to 
detect the nonlinear noise' at all (as it is signal-dependent and hard to 
detect), but instead rely on physics itself to do the cancellation for us," 
he says.

The researchers used two streams of light instead of the usual one when 
both signals are combined at the endpoint of a transmission, any "noise" 
created by interference within the optical fiber itself can be identified 
and canceled out, dramatically reducing signal loss and improving 
efficiency.

This works, according to the article, because these twin signals create 
distortions that are "essentially anti-correlated," meaning that, when 
combined, they should nullify each other.

"The idea came up during our study on the benefit of coherent superposition 
for improving the fiber transmission performance of an optical signal," 
says Liu, "particularly when the time-reversal picture of phase conjugation 
went into my mind: why don't we superimpose two phase-conjugated (twin) 
waves to see if their nonlinear distortions cancel each other automatically 
by nature? And it worked, especially when a symmetry condition is 
satisfied. The rest is now known."

Citing colleagues Peter Winzer and Andrew Chraplyvy, Liu said technology 
that takes advantage of this technique could become available within as 
little as three years' time.
--

Cheers,
Stephen

Message sent using MelbPC WebMail Server