Slashdot Mirror
Facebook Achieves 20Gbps Data Rate Over MMW Radio Spectrum (thestack.com)
An anonymous reader writes: Facebook's Connectivity Lab has announced that it has achieved data transmission rates of 20Gbps over the millimetre-wave (MMW) section of the radio spectrum; however, the transceiving stations need to be incredibly tightly calibrated to each other, with the team describing the margin for error as equivalent to 'a baseball pitcher aiming for a strike zone the size of a quarter'.
That's fantastic they can violate your privacy even faster now.
"That's the way to do it" - Punch
To be fair, as many people play baseball as snooker and snooker has a world championship too.
"That's the way to do it" - Punch
The article does not spell this out, but I lived in Woodland Hills so I know from looking at the map that the path between the receiver and transmitter is not even remotely line-of-sight. That's up over a very tall ridge and then back down onto the floor of the san fernando valley.
margin for error as equivalent to 'a baseball pitcher aiming for a strike zone the size of a quarter'.
For fixed-point unobstructed line-of-sight this isn't so difficult, especially if you use lasers to do your initial calibration.
The problems happen if you have obstructions (trees, rain) or movement (buildings move in the wind, so top-of-skyscraper to top-of-skyscraper would be hard to maintain). But for indoor use or near-ground-level use for tens or hundreds of meters, this might work. If you have a way of keeping calibration despite movement, these limits may be relaxed.
I can see this as a possible way to connect one end of a long factory to the other end without having to run wires. This assumes the places where you put the transmit- and receive antennas isn't subject to vibrations from the factory, of course.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Why not try using a sport the "world" plays - a world series that is played by one country...
Two countries participate in the World Series of Baseball, technically. Both Canada and the US have teams. But I would imagine that the best players in the world play in Major League Baseball, I think it's the highest paying league.
It's well understood that compression efficiency is inversely correlated with the amount of information contained in the data string. Since we're talking about Facebook, the information content is basically nil.
#DeleteChrome
Don't worry: it's still an painfully useless analogy if you are familiar with baseball. You really aren't missing anything.
In 2016 there were 6 countries represented. (and that's counting England, Scotland and Wales separately instead of under the banner of United Kingdom). There's only 2 countries represented by teams in the world series, but players come from 23 different countries.
I'm actually not disagreeing with the notion that "the world series" is a bit of a laugh as a name, but it is represented by people from a lot of countries, and to be fair, no team from outside North America would be competitive.
(there again there's probably more than 23 countries represented in the Premier League in England and we don't call it the World Premier League... yet).
Another one that amuses me is the Kabaddi World Cup. They actually have a couple dozen countries competing, but considering only two countries in the world actually play the sport seriously (and most of the competitors in foreign teams are ex-pats or children of ex-pats from India) it's a bit of a laugh.
"That's the way to do it" - Punch
Could we have a bad car analogy instead, that might be more understandable.
Could we have a bad car analogy instead, that might be more understandable.
The margin of error is smaller than the likelihood of actually getting a car to crank on the first try when being pursued by a serial killer.
"That's the way to do it" - Punch
If they were using a one-meter-plus parabolic antenna that had to be aimed to an accuracy of 0.07 degrees, and since they pointed out that this part of the spectrum is blocked by rain and fog, I'm going to say "totally". Millimeter-wave gets attenuated by just about anything. If the AC below is correct that there was a ridge in the way, then they must have been using really tall towers -- or, as hinted in the article, an airborne target.
oh yeah I see, a wifi system like that would be really stressful to operate
It's like aiming a car at the library of Congress and hitting only books by Ayn Rand.
Check out my sci-fi/humor trilogy at PatriotsBooks.
Please tell us more, O Expert AC, about the circuitry you use to generate an array of phase-controlled mm-wave signals.
Conveniently, they left out the effective radiated power (ERP) needed to get microwave radiation in the 30-300GHz range to a receiving point at the distance mentioned in the article of 13km.
If they were using a one-meter-plus parabolic antenna that had to be aimed to an accuracy of 0.07 degrees, and since they pointed out that this part of the spectrum is blocked by rain and fog, I'm going to say "totally". Millimeter-wave gets attenuated by just about anything. If the AC below is correct that there was a ridge in the way, then they must have been using really tall towers -- or, as hinted in the article, an airborne target.
I can't begin to imagine the ERP needed for sustained propagation in that frequency range, even with the +1m parabolic! Yow. But hey. Fast.
You don't have to worry about the wind drift. The Pringles can is able to squeeze the signal real tight.
“He’s not deformed, he’s just drunk!”
Still not as fast as an IR free-space link, and I'm struggling to see what other advantage it has. It's still blocked by atmospheric water or precipitation. It might be harder for a bird to block it by flying through the beam, but if you expand a laser beam's diameter to 1 m, no one bird's going to block that, either. And I can't imagine you'd need anywhere near 100 watts to get IR across that distance.
Ford claims making breakthrough in petroleum refining processes
love is just extroverted narcissism
'a baseball pitcher aiming for a strike zone the size of a quarter'
These comparison are really useful when have never seen a baseball field and only have an extremely vague idea of how far a pitcher is standing from the strike zone.
To be honest, I don't actually know how big an olympic-sized swimming pool is either
1/32" radio or some other absurd imperial unit?
Cost free eBook I read (by iBook/Kobo/Amazon/ObookO/Gutenberg etc.): "The Green Odyssey" by Philip Jose Farmer.
Ah, QA, I didn't recognize you without the space/3D-printing hat. Tell me, do you maintain a list of everyone who's ever criticized one of your posts? Because, while I get tired of your shtick pretty frequently, I don't think I've posted in response to you (or anyone else) that frequently.
In any event, the presentation you linked does show considerably better technology than I was aware of. (For anyone who couldn't already tell, no, this isn't my field.) And yet -- look at the performance graphs scattered throughout the deck. They seem to show a pretty messy 5-15db lobe being steered. Compare that to 50+db gain of a 1-meter parabolic antenna at 50GHz. Facebook is pushing 100W through that transmitter (although admittedly there's no indication how much of that makes it to the radiator). How do you suppose the steerable array system would fare if you fed it a megawatt?
(BTW, snark aside, if you are an antenna engineer, I'd love to see some actual analysis around this.)
Don't worry: it's still an painfully useless analogy if you are familiar with baseball. You really aren't missing anything.
Major League Baseball (MLB) rule 2.04 specifies that " the distance between the pitcher’s plate and home base (the rear point of home plate) shall be 60 feet, 6 inches" (18.44 meters).
MLB rule 3.01 specifies that the ball be "not less than nine nor more than 9 1/4 inches in circumference". That is 72.64–74.68 mm in diameter, let's split it down the middle and use 73.66 mm.
The MLB definition of a strike includes "a legal pitch when so called by the umpire, which: [...] is not struck at, if any part of the ball passes through any part of the strike zone".
A United States quarter is 24.26 mm in diameter.
Thus, the difficulty of aligning the transceivers is similar to the difficulty of throwing a ball of 73.66 mm diameter through a hole of 97.88 mm diameter from a distance of 18.44 meters.
There, now anyone in the world should have a good idea of how difficult it is. ;-)
heck it's easy if the "pitcher" is a computer controlled servo system that keeps the beam on point.
as equivalent to 'a baseball pitcher aiming for a strike zone the size of a quarter'.
I don't think that's American enough. Can we get it in terms of Statue of Liberties per acre of Freedom?