German Scientists' Visible Light Network Hits 3Gbps

Mark.JUK writes "Scientists working at Berlin's Fraunhofer Heinrich Hertz Institute have developed new components that can turn standard 'off-the-shelf' LED room lights into an Optical Wireless Local Area Network (OWLAN) that delivers data transmission rates of up to 3Gbps. The new kit is an extension of HHI's earlier work, which in 2011 delivered the first 800Mbps capable network using ordinary flashing LED lights. Since then the kit has been improved to achieve a transmission rate of 1Gbps per single light frequency (basic LEDs usually use up to three light frequencies) and the operating bandwidth has been pushed to 180MHz from 30MHz."

Comments

sounds overly optimistic

[Anonymous Coward]

"basic LEDs usually use up to three light frequencies" is BS. Nobody uses RGB for room lighting - color reproduction is not good enough. You use blue LEDs + photoluminescent phosphors. I wonder whether they can also mudulate the phosphors at 1 Gbps, but I doubt so.

Re:sounds overly optimistic

[K. S. Kyosuke]

Nobody uses RGB for room lighting - color reproduction is not good enough.

They probably wouldn't be the sole light source in the room - I imagine that this is just a bit of added value.

Re:sounds overly optimistic

[russotto]

I wonder whether they can also mudulate the phosphors at 1 Gbps, but I doubt so.

You wouldn't need to; enough of the original LED color gets through the phosphors to detect.

This isn't exactly a new idea; it's been known for years that you can read the data from an old-style modem's Tx and Rx LEDs from across the room.

Re:sounds overly optimistic

[Khyber]

"Nobody uses RGB for room lighting"

You're wrong. In fact, not only can an RGB diode produce great white light, we have diode packages that can essentially cover the entire visible spectrum and thus create any CCT known with greater efficiencies than a white diode, which, again, you're wrong - it's a UV diode with a phosphor on it, not a blue diode.

Re:sounds overly optimistic

[Teun]

So?

Where did you read they want to use the regular/primary room lights for this sort of communication?

Re:sounds overly optimistic

[Teun]

Using standard LED's is still a way off using a room's main lights.

The way I understand it is they use off the shelf visible-light LED's in stead of the for communication more regular UV or IR version.
Cool and one day it might end up in a room's primary illumination, meaning there's no network during daylight hours :)

Re:sounds overly optimistic

[SternisheFan]

thanks, I had confused 'room lights' with 'standard LEDs', I guess. :^\

Re:sounds overly optimistic

[JRIsidore]

... it's a UV diode with a phosphor on it, not a blue diode.

Nope. Just look at the spectrum of some white LEDs, they clearly peak around 450 nm plus what the phosphor delivers. UV is very problematic as it quickly degrades the plastic optics which are predominantly used with LEDs. Plus, you would only get the yellow light from the phosphor, not white light. It's the mixture of blue and yellow that's necessary where the ratio determines the correlated color temperature.
E.g.: http://www.cree.com/led-components-and-modules/products/xlamp/discrete-directional/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/Data%20and%20Binning/XLampXPG2.pdf

Re:sounds overly optimistic

[Khyber]

I do LED work for a living. I work very closely with Cree, Nichia, etc.

All good high-efficiency white diodes are UVB diodes with a triple-component or more phosphor layer, and possibly with a ceramic recombination package design. You start with UV, add a blue phosphor base on top of that, then add your amber and red phosphors on top of that, then try to use the packaging to redirect light that scatters back through the phosphors for maximum output.

"UV is very problematic as it quickly degrades the plastic optics which are predominantly used with LEDs"

I have raw UV diodes in most of my growing panels. They all have plastic covers. Not one single problem. Teflon is quite UV resistant, and is used quite often in this particular application, mainly because of the localized heat put off by an LED plus the UV emissions.

"Plus, you would only get the yellow light from the phosphor, not white light."

Wrong again, as per above. I'm holding engineering samples of MK-R, 200 lumen/w @ 7000K CCT and a 93 CRI. You put it through a diffraction grating, and you see EVERYTHING. From ~370nm up to 700, no line breaks or indications of narrow-range phosphors. In fact, most newer white LEDs have essentially blackbody output, you can't tell a difference under a spectrometer between incandescent and LED, except that the LED is far brighter (with the exception around 680-700nm) and misses most of the IR/NIR range. Here, have a picture.

Also, note your chosen datasheet only goes as low as 380nm. Not even close to the UVB range, which is where white diodes begin their life. Go put one of your XPG2 over a sheet with yellow highlighter on it or one of those green USPS tracking confirmation labels. Watch it fluoresce like mad. Even most blue diodes start with a UVB base, including the ones used in the growing lights I design. This isn't Near-UV, here. This is full UV.

And its the mixture of green and blue and red that makes CCT. RGB. Yellow is not included in the official CIE 1960 color space as a primary source.

Harold Haas - links

[SternisheFan]

Harald Haas: Communications technology innovator: Harald Haas is the pioneer behind a new type of light bulb that can communicate as well as illuminate – access the Internet using light instead of radio waves.

TedTalks - Why you should listen to him:

Imagine using your car headlights to transmit data ... or surfing the web safely on a plane, tethered only by a line of sight. Harald Haas is working on it. A professor of engineering at Edinburgh University, Haas has long been studying ways to communicate electronic data signals, designing modulation techniques that pack more data onto existing networks. But his latest work leaps beyond wires and radio waves to transmit data via an LED bulb that glows and darkens faster than the human eye can see.

The system, which he's calling D-Light, uses a mathematical trick called OFDM (orthogonal frequency division multiplexing), which allows it to vary the intensity of the LED's output at a very fast rate, invisible to the human eye (for the eye, the bulb would simply be on and providing light). The signal can be picked up by simple receivers. As of now, Haas is reporting data rates of up to 10 MBit/s per second (faster than a typical broadband connection), and 100 MBit/s by the end of this year and possibly up to 1 GB in the future.

He says: "It should be so cheap that it’s everywhere. Using the visible light spectrum, which comes for free, you can piggy-back existing wireless services on the back of lighting equipment."

"As well as revolutionising internet reception, it would put an end to the potentially harmful electromagnetic pollution emitted by wireless internet routers and has raised the prospect of ubiquitous wireless access, transmitted through streetlights." Herald Scotland

http://www.ted.com/speakers/harald_haas.html

Here is the TED talk video:

http://www.ted.com/talks/harald_haas_wireless_data_from_every_light_bulb.html

Re:Harold Haas - links

[sanman2]

Is this a secure mode of communication? Or are you going to need some kind of Lighting Encryption Protocol?

Re:Harold Haas - links

[ceoyoyo]

"As well as revolutionising internet reception, it would put an end to the potentially harmful electromagnetic pollution emitted by wireless internet routers and has raised the prospect of ubiquitous wireless access, transmitted through streetlights." Herald Scotland

It's too bad you had to throw that one in. It's pretty funny though: "let's use flashing lights instead of electromagnetic radiation!"

Re:Harold Haas - links

[CastrTroy]

Any wireless network is going to need encryption. This is the reason why we have WPA for our radio frequency wireless networks. You could probably use the exact same security protocols as I'm pretty sure they don't depend on the medium you are transferring over.

Re:Harold Haas - links

[sanman2]

Well, he said "electromagnetic pollution" and not "radiation" - so I think he means that just lighting is better than lighting+wireless.

Re:Harold Haas - links

[Teun]

Ah!

I've just put in my patent application for * to include 'via light'.

Next week you can find me on my personal tropical island.

Re:Harold Haas - links

[Ol Olsoc]

Well, he said "electromagnetic pollution" and not "radiation" - so I think he means that just lighting is better than lighting+wireless.

And still digital people seem to think that bandwidth is infinite.

Modulating a light beam is incredibly trivial. The only limitation for frequency or bandwidth is the amount of time it takes to turn the light off, then back on. I could have somethign running in about ten minutes at the workbench. A reciever is a little more intricate, but still this is something that middle school students with a little electronics aptitude could figure out. Heck, Infrared Television remote controls are doing this now.

The questions I would have are say, in the case of the automobile, how are you going to send and recieve that data? Possibly arrays of recievers along the roadsideI'm not quite certain how well that is going to work with crowded highways. Also from a security standpoint, it is so terribly easy to jam.

I would see the main use of LED modulation in autos as a way for Law enforcement to get into the car's working's, ID'ing the car registration, perhaps the driver, perhaps disabling the vehicle.

For in home or in room use, there would be some potential. An IR LED in a corner of the room with whatever device is to be networked is very promising. Eliminate wiring, perhaps a little more secure, as long as a balance between IR power and reception can be maintained - again, think infrared remotes and their issues. Maybe keep the windows closed?

Visible Light Wireless Network

[Anonymous Coward]

Epileptic seizures sure make the download time breeze by.

Re:Visible Light Wireless Network

[ColdWetDog]

When you first heard one of the "a horse walks into a bar" jokes, did you reply by citing a law the bans horses from bars and then explain that the premise of the joke was simply untenable?

I hope that your grasp of sarcasm is better than your grasp of humor.

You see, this is a problem with kids these days. They didn't grow up with "Mr. Ed" or "Green Acres" (or "I Dream of Jeanie" for that matter, but I digress). The subtle twists of cross species humor are just lost on them.

And no, Dick Cheney doesn't count.

Re:Visible Light Wireless Network

[chill]

Dick Cheney? Being undead isn't a different species.

Re:Visible Light Wireless Network

[Arancaytar]

Any flicker at 100Hz or faster is imperceptible to humans (and even that's pretty generous; the actual threshold is lower). This is 30-180MHz; a million times as fast.

Re:Visible Light Wireless Network

[Ol Olsoc]

You see, this is a problem with kids these days. They didn't grow up with "Mr. Ed" or "Green Acres" (or "I Dream of Jeanie" for that matter, but I digress). The subtle twists of cross species humor are just lost on them.

And no, Dick Cheney doesn't count.

Anyone who did not grow up with "I dream of Jeanie" was terribly terribly impoverished.

I suppose the show is terribly un-PC these days, but did I did indeed dream of her.

Then again, I also believe that Green Acres was the height of Western Civilization.

Link to article

[Vario]

Unfortunately the press release is a little short on details. Here is the link to the actual article (paywalled):

"1.25 Gbit/s Visible Light WDM Link based on DMT Modulation of a Single RGB LED Luminary", opticsinfobase.org

But this is only half the problem

[Antony T Curtis]

Fast download rates, okay. But what about the return/upload path?

Re:But this is only half the problem

[ceoyoyo]

Most computers have some LEDs in them. iPads would need a dongle or hardware revision.

Re:But this is only half the problem

[Antony T Curtis]

But detecting that light from the flood of all the other light is a problem.

Re:But this is only half the problem

[uninformedLuddite]

LOL. He said dongle.

Re:But this is only half the problem

[ceoyoyo]

I can see the LEDs on my notebook perfectly well in a lit room. You probably wouldn't be able to coax 3 Gbps out of them, but most people don't use as much upstream bandwidth as down. Exceptions are mostly wired machines anyway - gamers and servers. You could also use IR LEDs for the upstream channel if you wanted to have a full duplex network.

Scientist?

[mark_reh]

Maybe engineer is a better term...

Re:Scientist?

[Anonymous Coward]

Maybe engineer is a better term...

This distinction is lost to the Germans.
Here you can study something called engineering-sciences...

Pre-existing technology

[SternisheFan]

I've been using a pair of Infra-Red Wireless headphones ($40 from Radio Shack) for some time now and the IR tech is impressive. While inside the room where the transmitter is there's really no interruption of the signal at all (it helps the transmission a lot when your walls/ ceiling are painted white to bounce the light off of). This sounds like a re-application of this pre-existing technology, and I'm not sure why it hasn't become mainstream for transmitting computer data already.

Re:Pre-existing technology

[ColdWetDog]

Maybe because he's pushing several orders of magnitude more data through the system than your 20 khz headphones?

Size matters.

Re:Pre-existing technology

[SternisheFan]

Maybe because he's pushing several orders of magnitude more data through the system than your 20 khz headphones?

Size matters.

A-a-a-nd, that's what she said, thanks for that opening line :-)

Okay, I read from the pdf page that it's expected to be available 3 years down the road, either in USB stick form or built-in smartphone sensors.

http://www.hhi.fraunhofer.de/media/downloads.html

Re:Pre-existing technology

[Forever Wondering]

IIRC, IBM, also, published a white paper on a system similar to Fraunhofer's about 10 years ago [to head off broad patent claims] based on research they had done.

Re:Who gives a flying monkey's??

[ColdWetDog]

All of that is Obama's fault.

Re:Who gives a flying monkey's??

[Teun]

You are not much into history.

A couple of centuries ago this guy Mohammed wanted to see the mountain and when it didn't show up on his doorstep instead he went to the mountain.
And would your present smart phone have the same speed and options as the one of 10 years ago it would probably last a lot longer now.

"Basicode"

[Anonymous Coward]

Back in the 80s, a radio station used -as part of a show on computers- part of the air time to distribute software; you just recorded the show on a regular tape deck, then used a BasiCode-decoder software and cassette recorder on your computer to load it.

So all "we" now need to do is to hack in the LED-based street lights on highways, and we can pump the latest software to car-based systems.

But seriously, you might be able to distribute low-data-rate stuff like traffic information,etc. into the lights to on-board systems using this technology one day.

How does it work through walls?

[__aaltlg1547]

Seriously, I can't see a practical application for this in combination with room lighting. And in the typical multipath light environment of a room that people live and work in, your speeds are going to be a lot less than what they measured under optimal conditions. One advantage though: only adding a conventional telescope, you could establish point to point links through open air over miles without breaking any FCC (or agency in your-country-of-choice) emission rules.

Re:How does it work through walls?

[Dthief]

FTA: The cheap LEDs, which could for example be placed on the ceiling or in room lights and tend to have coverage of around 10 meters, essentially blink on and off extremely fast to transmit the data (not visible to your naked eye). This would make it extremely useful for short range and high-speed networks that may also require something more secure than wifi (i.e. light doesn’t travel so well through solid walls etc.). So it IS the room lighting, and yes, it is not meant for long range wireless. But you could link everything in a room to it.

Re:How does it work through walls?

[brunes69]

There are a lot of companies and organizations where this would be considered a benefit since they would no longer need to construct special buildings to block wireless from leaving high-security rooms / the building.

Re:Who gives a flying monkey's??

[__aaltlg1547]

Or by pouring vodka in it. Remember that? Booze-Fueled Gadgets

PARCTAB

[stenvar]

The original PARCTAB, basically the first computer to roughly look and work like a modern touch screen device, used networking based on ceiling-mounted LEDs. A paper describing the system is here. Many systems used IrDA communications after that. Of course, it's probably been a lot of engineering work increasing the speed of the system, but it's not a fundamentally new idea, just the evolution of old technology.

Re:PARCTAB

[tlhIngan]

Many systems used IrDA communications after that. Of course, it's probably been a lot of engineering work increasing the speed of the system, but it's not a fundamentally new idea, just the evolution of old technology.

802.11 also includes an infrared PHY layer in the standard - sure it was only 1-2Mbps, but it was there. This was the original 802.11 standard, but it was there as an alternative to IrDA.

Congratulations, Dear all at HHI!

[udippel]

HHI used to be the world championship in optical signal transmission beating their own records as early as the late 1970 and early 1980. I myself had the honour to work there, at that time, though not in optical transmission systems. The time spend there has always been a great and endearing reminiscence.
I am proud of you, guys and girls! Congratulations!
(I really wonder if anyone from those days is still there!?)

Re:Who gives a flying monkey's??

[SuperKendall]

It has a much higher chance of reaching the market than any project from grad students working with large quantities of alcohol.

Someone should really go check on those guys...

Re:Who gives a flying monkey's??

[Bengie]

We were limited to 33.6Kb because modems had to interface with one of the many T1 channels. DSL by-passes the T1 and can use a more modern Layer 1.

Cool, but...

[stripes]

...can anyone come up with a use for this that existing WiFi doesn't already cover? It isn't more range, and I'm not sure it is usefully less range. If you are worried about eavesdroppers on the network you need light tight rooms, but if you want to set up a whole house network you need to have repeaters for each room.

This seems more like an answer in search of a problem. Sometimes that means we will find a problem we didn't understand we had, and sometimes this turns out to be the technology equivalent of the big kitchen junk drawer full of bits that almost never get used.

Woah.

[Anarchduke]

That's a lot of German porn.

Re:Who gives a flying monkey's??

[uninformedLuddite]

Someone should really go check on those guys...

I suspect they are on their private yachts. They didn't actually make anything but the IPO was a hoot.

Asymmectric networks would be very useful.

[ksbraunsdorf]

At the library I want to see a list of all the titles in print from my favorite authors, I just use the local WiFi to get the data. For larger downloads I ask on the WiFi, but get the data over the visible light network. So I can see the text of all those books, DRM allowing. Or watch a lecture on the Great Bustard. At airports, my PDA/phone gets all the flight updates on an endless loop, via the visible light network. At Home Depot I'm offered product information and How To videos. I'd love to see the view from the cockpit in real-time while I was flying. If we build really high capacity broadcast networks (like the over-the-air TV used to be), then we'll find uses for them we've never thought about at all. This may even make a computer useful in a class room. I don't believe most of this requires encryption. Mostly an asymmetric network gets us video and large data requests over a cheap, local, and very limited range network. If you want encryption for small slices of data, us the WiFi to do a key exchange.