They're likely not using something too exotic: 10Gbps free space optical links have been available commercially for a few years now (see products by AOptix). The tricky part is pointing them and getting the last bit of power on target. If they're actually forming links at 100's of km separation they're already on par with what APL was doing just 2 or 3 years ago [1]. Impressive that they not only have the data terminal to develop but have had the additional complication of building a new airframe as well.
Having worked on the pointing/tracking/acquisition side of this problem that's the part that's truly impressive to me. Sub-microradian pointing is not trivial. Their "red beam following green beam" shots in the video must be included for a bit of a laugh...
Agreed, this is very cool. How do free space optical links work in the rain though? Shouldn't be a problem for the link between aircrafts as they fly above the weather but might be for aircraft-ground link if that's optical.
Hope FB will open source parts of this! I know NASA has been looking at optical links for high bandwidth space communication, this accurate tracking would really help
There are a couple effects from rain: absorption ("db per km") and scattering (bigger constant in front of your r^2 term), neither are good for FSO. Fog (and since this is airborne, clouds) is the real killer. Imagine taking your nice .1 degree wide laser beam and pointing it into one of those frosted street lamp bulbs so that it now points in every direction, then add to that the attenuation from being absorbed by the droplets. Now, in FB's case, they're doing something pretty smart to mitigate the water problem. Using RF for the ground-air links avoids most of the problems with scattering and absorbtion depending on droplet size and the band you're transmitting in. Hard to beat the bandwidth of FSO, so that's a great option for above the clouds. 60,000 feet should be above most clouds - anything that tall and you'll be wanting to get out of the way in any case.
The LCRD stuff that NASA is doing is indeed impressive. I think that their pointing problem is a few orders of magnitude harder than what FB is dealing with though - if anything we'll see technology transfer in the other direction.
Does anyone know how they're achieving the 10gbps data rate? I'm wondering if they're doing something like vortex beam forming... http://www.gizmag.com/twisted-light-beam-data-transmission-v... and http://www.sciencedirect.com/science/article/pii/S0030401809... (Although this article is about using a physical fiber).