'Twisted light' carries terabits

[Robert Seastrom]

On Jun 27, 2012, at 1:41 PM, André Kesteloot wrote:

> ** 'Twisted light' carries terabits **
> Researchers show off a novel way of fitting more data on to light waves by "twisting" them - and clock beams carrying 2.5 terabits of data per second.
> < http://www.bbc.co.uk/news/science-environment-18551284 >

Similar article to the one yesterday.  Doing it with point to point free-space optics or on fiber is a heck of a lot more practical in my opinion than doing it with some kind of "wifi-like" technology.  I blame shoddy reporting in the previous article.

We already do cross-polarized light on fiber (2-way).  8-way would seem to just be a refinement of the old trick.  Cross-polarization isolation is in theory infinite, but in the real world it seems to be 20 to 40 dB for RF stuff.  Not sure about glass; I'd expect it to be at the high end or maybe higher.  Question is how well does that work when he cross-pol angle is < 90 degrees?  I wonder how much effective throughput adding the twist gets you.

-r