Researchers Set World Record Wireless Data Transmission Rate of 6 GB/Sec Over 37 KM (sciencedaily.com)

An anonymous reader quotes a report from Science Daily: Transmitting the contents of a conventional DVD in under ten seconds by radio transmission is incredibly fast -- and a new world record in wireless data transmission. With a data rate of 6 Gigabit per second over a distance of 37 kilometers, a collaborative project with the participation of researchers from the University of Stuttgart and the Fraunhofer Institute for Applied Solid State Physics IAF exceeded the state of the art by a factor of 10. The extremely high data rates of 6 Gbit/s was achieved by the group through efficient transmitters and receivers at a radio frequency of 71-76 GHz in the so-called E band, regulated for terrestrial and satellite broadcasting. The circuits are based on two innovative transistor technologies developed and manufactured by the project partner Fraunhofer IAF. In the transmitter the broadband signals are amplified to a comparatively high transmission power of up to 1 W with the help of power amplifiers on the basis of the novel compound semiconductor gallium-nitride. A highly directive parabolic antenna emits the signals. Built into the receiver are low-noise amplifiers on the basis of high-speed transistors using indium-gallium-arsenide-semiconductor layers with very high electron mobility. They ensure the detection of the weak signals at high distance. The transmission of high quantities of data by radio over large distances serves a high number of important application areas: the next generation of satellite communication requires an ever-increasing data offload from earth observation satellites down to earth. Supplying the rural area and remote regions with fast Internet is possible as shown in the trial. Earlier this year, engineers at the University of Illinois were able to set a record for fiber-optic data transmission, transmitting 57Gbps of error-free data at room temperature.

Comments

Units

[Anonymous Coward]

GB/sec != Gb/sec -PCP

Re: Units

[johnsmithperson123]

Yup, maybe ~700MB/s (megabytes a second.) But "does transmitting a whole CD in a single second" sound as good?

Re:Units

[SpelledBackwards]

And over a distance of 37 Kelvin-Mega!

Science really does push the boundaries of what I thought possible.

Re: Units

[Anonymous Coward]

Slashdot is not run by nerds anymore, they don't know the difference between bytes and bits.

The editor should be fired immediately. But, he won't, because this site is a joke now.

Re:Units

[MightyMartian]

Look, asshole, I happen to know someone who made the Kessel Run in 12 parsecs, so don't getting all unit-Nazi around here!

Re: Units

[MightyMartian]

And yet here you are...

Re: Units

[Anonymous Coward]

Boooooooooooo

Re:Units

[Anonymous Coward]

What's an order of magnitude between strangers?

Re:Units

[rossdee]

and are those Gigabits (1,000,000,000) or Gigibits (1,073,741,824)

Re:Units

[Ol Olsoc]

Look, asshole, I happen to know someone who made the Kessel Run in 12 parsecs, so don't getting all unit-Nazi around here!

Yeh, and he shot first.

Re:Units

[alvinrod]

Depending on what the Kessel Run actually is, that statement might make sense. If it were something that required a lot of tricky navigation, a shorter total distance traveled could well be an indication of skill as opposed to someone who was less skilled and had to take a longer path.

It's probably not that, but it's salvageable if you try real hard.

Re:Units

[Anonymous Coward]

Stone cold killer, that one. Stone. Cold. -PCP

Re:Units

[Anonymous Coward]

I thought it was 14.

Re:Units

[Anonymous Coward]

He didn't shoot first, he shot only.

Re:Units

[Anonymous Coward]

That's tenuous retcon garbage to try to cover for George Lucas's lack of knowledge.

Re:Units

[Aighearach]

How many Cowboyneals per hour is it?

Re:Units

[hcs_$reboot]

And it's 37 km, not 38. Speed of light is so slow that 38 km would have made a huge difference.

Re:Units

[Anonymous Coward]

God, I hate idiots like you. Someone does NOT have to shoot 2nd for you to shoot 1st.

Re:Units

[omnichad]

If you're going to be pedantic, get it right - gibibit

Re:Units

[omnichad]

And now I realized I didn't capitalize (implodes in a rage of pedantry)

Re:Units

[Shatrat]

Gigabits. In telecom we don't measure line rate in bytes. Bytes only come into discussion when talking about framing.

Re: Units

[Anonymous Coward]

But "does transmitting a whole CD in a single second" sound as good?

Not to the MAFIAA.

Re: Units

[Anonymous Coward]

You just gave me a stomachache

Re: Units

[Anonymous Coward]

Nope. It's a social justice site now.

They can perform privilege calculus in their heads (ie. make stuff up), but can't tell the difference between bits, bytes. They think RAM is gendered and therefore sexist.

Re:Units

[Anonymous Coward]

It's probably not that, but it's salvageable if you try real hard.

Imagine a "shuttle run" type race where your carrying capacity is being measured, kind of like running Moonshine. Boasting that completing the Kessel Run in 12 parsecs would imply that the Millennium Falcon has a high carrying capacity.

Captcha: beholds

Re:Units

[Anonymous Coward]

Uh, yes, they do. If you run around a track by yourself, you wouldn't finish first because you weren't racing against anyone else. You would be the ONLY finisher, shit for brains.

Re:Units

[Anonymous Coward]

Look, asshole, I happen to know someone who made the Kessel Run in 12 parsecs, so don't getting all unit-Nazi around here!

On the face of it, that measurement doesn't make sense, however, with a bit of background knowledge, it makes perfect sense. The Kessel run is a route through a area full of black holes, the more precise your navigation, the shorter distance you have to go to make it through the area (you can go closer to the blackholes without running into them).
But, in saying that, with decent sublight engines, you could probably cut the distance more but it would take you a lot longer to get through so...

Re:Units

[Ol Olsoc]

This is why we can't have anything nice.

Re: Units

[Coren22]

It is also a microaggression to point out the misuse of units!

Re:Units

[lsatenstein]

If that can be commercialized, then AT&T and the others will have 10x bandwidth to sell. No, you won't get a discount. Never give a captive client (sucker) any even break.

Re:Units

[Anonymous Coward]

How many meters in a Gibimeter?

B or b?

[Anonymous Coward]

Pedantic? Yep!

Necessary on a site like this? Yep!

I almost shit myself when I read 6 GB/s, that's fucking amazing...

6 Gbps is slightly different, like, 8 times as different.

Re:B or b?

[PIBM]

If only that was the only unit issue. Still, that one was so bad that I was initially interested, until I read through the summary and saw the different notation. Too bad, lets revisit in 10 years.

Re:B or b?

[ShanghaiBill]

6 Gbps is slightly different, like, 8 times as different.

If we are going to get pedantic, let's go all the way: The bit rate (little b) generally includes protocol overhead (framing, ECC, handshaking, etc), while the Byte rate (big B) generally refers to only actual data. So the difference is not a factor of 8, but usually around 10.

Re:B or b?

[TechyImmigrant]

Pedantic? Yep!
Necessary technical detail? Yep!

How necessary? 8 times as necessary!

6 GB/s is fucking awesome!

6Gb/s, not as impressive.

But if Comcast was selling it as a service, they would cap you after 3 seconds of full bandwidth use.

Re:B or b?

[Anonymous Coward]

If you don't think wireless link of 6Gb/sec over 37 kilometer link isn't impressive, then you need your head checked. Holy fuck that is impressive.

Re:B or b?

[MightyMartian]

So by the time this is done, it will have about the same transfer rate as a USB2 flash drive.

Re:B or b?

[Lije Baley]

Ah, thanks, I thought DVDs were getting bigger there for a second...or 10 seconds I guess.

Re:B or b?

[Anonymous Coward]

4.7E9 / 0.6E9 = 7.8 seconds. Of course with error correction will end up being more like 9.36 seconds.

Re:B or b?

[OrangeTide]

Unfortunately capital-B meaning bytes is not universal in all industries. In telecom I can't think of any time you would measure a data rate in bytes, it's usually the bit rate and sometimes the symbol rate.

Re:B or b?

[Anonymous Coward]

No, 6Gb/sec DATA RATE is USB 3.0

Re:B or b?

[Anonymous Coward]

They didn't say it's not impressive, they said it's not AS impressive as the typo would indicate.

Re: B or b?

[Midnight Thunder]

If you consider most people are probably getting a speed nearer 25mb/s for their DSL and 100-200mb/s for their home wifi, then 6Gb/s is still pretty impressive over that distance, even if it is not GB/s. This could be really good for a mesh network, where you could be sharing videos with neighbours in a rural community.

Re: B or b?

[afidel]

The units are all messed up, the record has to be in bytes because we have relatively inexpensive 100Gbps commercial systems today using 4x 25Gbps as the carrier, there's no way the best lab system is only 2x the cots rate.

Re: B or b?

[Sique]

Really? 25 Gbit/sec wireless over 37 km?

There is quite a difference between sending data via fiber optics over 10,000 feet or wireless over more than 20 miles.

Re:B or b?

[AmiMoJo]

6Gb/sec could be very useful for providing broadband to rural areas.

Re:B or b?

[Anonymous Coward]

Psst! 6Gbps is theoretical max for SATA III. 5Gbps is theoretical max for USB 3.0.

Re: B or b?

[TechyImmigrant]

It's very impressive indeed. Wireless point to point at that frequency is magical. We used to think 5GHz was heady stuff when we were developing the 802.11 specs. That's p-mp, but still, the front end devices dictate what's possible.

Re:B or b?

[Coren22]

But, in telecom do they capitalize it?

Verizon doesn't:
http://www.verizon.com/home/fi...

AT&T doesn't:
https://www.att.com/internet/

So, who uses B when they should use b?

Re:B or b?

[Anonymous Coward]

ATM, T1 / DS-1 and other gear typically use MB next to the modular jacks. As they don't have space for "Bits" and for whatever reason didn't want to mix upper case and lower case letters (paying for upper case only fonts maybe?)

reaching back into ancient history 3com network cards use B and sometimes b.

As far as I know CCITT uses Mb/s universally in their documents. Which is far more compelling than your AT&T and Verizon examples.

Re:B or b?

[Coren22]

As far as I know CCITT uses Mb/s universally in their documents. Which is far more compelling than your AT&T and Verizon examples.

That was kind of the point I was making. I don't know about the T1 jacks you have seen, but I haven't ever seen one that was labeled with MB or frankly Mb. Everywhere I know of the telecoms industry understands the difference, and to avoid lawsuits they try to use Mb where it applies.

B or b?

[killfixx]

Pedantic? Yep!
Necessary technical detail? Yep!

How necessary? 8 times as necessary!

6 GB/s is fucking awesome!

6Gb/s, not as impressive.

MIMO

[Anonymous Coward]

Pretty obvious from the article that this is a single antenna. Wait until this is transferred to an ASIC and has multiple antennas. It could reach up to 24Gb/sec possibly. Of course won't have the same distance, but won't be surprised if this scales to over 20Gb/sec indoors. This is HUGE breakthrough. Congrats to all those involved with this!

Re:MIMO

[bosef1]

As the summary says (and no-one has contradicted), this experiment used a directional parabolic antenna to improve the performance of the data transmission. Along with parent, I am not really all that excited until they can get the same performance with either multiple antenna or an isotropic radiator to give 4-pi steradian coverage (or so). Having a point-to-point link is nice, but I can do that with lasers and microwave beams already.

Re:MIMO

[Anonymous Coward]

Parent, here. You definitely misread my statement. I'm very excited about this announcement. With MIMO, these guys may achieve indoor wireless rates sufficient for current VR headset technology. I didn't expect this to happen for maybe 5-10 years.

Re:MIMO

[Anonymous Coward]

A directional antenna like they describe at 71GHz is very, very directional. As is the receiver. VR headsets need nearly isotropic (need about a hemisphere or half a hemisphere) in order to keep up with motion tracking. These antennas likely are painting something the size of your eyeball. Not something that's going to keep lock for VR.

Re:MIMO

[Anonymous Coward]

Very directional at 37 kilometers... I only need 15 feet.

Re:MIMO

[Anonymous Coward]

They use very directional antennae for a 37 km link. Going from 37 km to 10s of m is a difference of 70 dB, and I don't think their antenna is anywhere near that directional (a 15 m parabolic dish at their frequency would have an ideal gain of ~60 dB). Also, they wouldn't be aiming at something the size of your eye, because at 70 GHz, a diffraction limited beam the size of your eye would need a couple meter diameter antenna just a few meters away to focus it down that much. And MIMO would easily be able to keep a lock on something in a volume of a couple meters, and that can give you back ~20 dB.

Re:MIMO

[Blaskowicz]

Look Wireless Displayport up : that's really something that seemingly competes with a VGA cable and such. At the same time it falls short for very high res at 90Hz and you would likely not get 100% bitrate all the time. There is ultrafast compression built into DP 1.3 that will certainly be useful. Might not cut it still. Then, GPUs can render some of the scene at partial resolution (recent nvidia PR), wireless VR would have to integrate that saving in the display chain. If that's not enough you'll need some higher frame by frame compression which will introduce latency, which is the VR devil.

I think it's close but using a cable would be safer. At least USB type C is there for that (run fast DP on the high bandwith portion not USB 3.1, use USB 2.0 for everything else)

Hey anyone else notice?

[im_thatoneguy]

Hey? Did anyone else notice the headline is wrong?

Re:Hey anyone else notice?

[Anonymous Coward]

This is Slashdot. A headline being wrong is in no way noteworthy. Many of them not even mistakes, but rather flat-out lies. Anything for the clicks!

Data plan ?

[Archfeld]

Great so now you can blow through your entire year of data in like 14 seconds of transfer, and AT&T can charge you like $10K in overages in less time than it takes to dial up support. T-Mobile will of course still offer unlimited data provided you don't actually transfer any because that will 'endanger' the stability of their network and you will be disconnected, plus incur premature contract severance fees. Cricket will of course offer a budget plan geared towards seniors who won't know any better. Tracfone will be illegal because it will require a DNA sample to buy a phone or any minutes...

Note : The above was intended for the purpose of satire and should not be taken literally...

what about 802.11ad?

[Anonymous Coward]

it goes as fast.. and given the right power mix and antennas, should be able to do the distance, too

802.11ad uses ordinary wifi frequency bands 2.4/5ghz plus a band in the 60ghz range (not far off these folks' 70ghz) .. and ordinary wifi can stretch to 25 miles and beyond with the proper equipment (and without pissing off the fcc).

Re:what about 802.11ad?

[Blaskowicz]

The spectrum used by it is unlicensed because it is absorbed by the air's oxygen, if you look it up it's one of the spikes oxygen has on a graph. Long distance might not work at all (there might be some slack at the upper or lower range).
If the link can survive rain or mist for something line of sight and across the street(s) we'll be lucky enough. I don't know. Good thing is fallback to 2.4 or 5.5 is pretty much built-in if that's of use.

No idea about the power use (mobile battery life). The immediate usefulness is shit tons of very local spectrum i.e. the conference room and the living room ; beaming 1080p60 and such (interactive desktop or display, not a video)

Re:what about 802.11ad?

[LeadSongDog]

It looks in the press release ( http://www.iaf.fraunhofer.de/e... ) that they've got about a 0.6m dish diameter, so at 4mm wavelength that's about 50dB gain and 0.5 degree beamwidth. Match that at the Rx end for 100dB total antenna gain. Use the full 1W Tx power they claim to have. Suddenly you've got a lot of room for loss in the channel.

Not the state of the art

[Anonymous Coward]

600Mbps is not the state of the art. There have been satellite hardware available 10 years ago that could transmit 600Mbps. Several satellites do that now. I think some now even do 800Mbps to 1Gbps - and have been doing that for at least a year or two.

Down, boy

[Anonymous Coward]

"Stuttgart and the Fraunhofer Institute" ? Wow, eurotrash is getting uppity. Anyway, I am sure it will be American company that actually makes shitload of money on this technology.

Bandwidth of a truck and all that...

[soramimicake]

6Gbps is about 2TB / hr.

Just put a 3TB HDD (or lots of DVDs) on a car and drive to the destination in half and hour, and you've achieved more than double the bandwidth :-P

Re:Bandwidth of a truck and all that...

[Anonymous Coward]

Only if you ignore the time it takes to read/write the HDD (or DVDs, or magnetic tapes, or whatever your storage medium of choice is). Short of SSDs, I'm not aware of a persistent storage format that won't bottleneck the truck.