Slashdot Mirror
1Gbps Wireless Network Made With Red and Green Laser Pointers
MrSeb writes "Back in the olden days, when WiFi and Bluetooth were just a glimmer in the eye of IEEE, another short-range wireless communications technology ruled supreme: Infrared Data Association, or IrDA for short. IrDA was awful; early versions were only capable of kilobit-per-second speeds, and only over a distance of a few feet. Trying to get my laptop and mobile phone to link up via IrDA was, to date, one of the worst tech experiences I've ever had. There's a lot to be said for light-based communications, though. For a start, visible (and invisible) light has a frequency of between 400 and 800THz (800 and 375nm), which is unlicensed spectrum worldwide. Second, in cases where you really don't want radio interference, such as hospitals, airplanes, and other sensitive environments, visible light communication (VLC), or free-space optical communication, is really rather desirable. Now researchers at the National Taipei University of Technology in Taiwan have transmitted data using lasers — not high-powered, laboratory-dwelling lasers; handheld, AAA-battery laser pointers. A red and green laser pointer were used, each transmitting a stream of data at 500Mbps, which is then multiplexed at the receiver for a grand total of 1Gbps."
> visible (and invisible) light has a frequency of between 400 and 800THz (800 and 375nm), which is unlicensed spectrum worldwide.
Well, that's good.
I thought IrDA was a famous Starcraft player...
This is old hat:
http://www.airlinx.com/products.cfm/product/19-0-0.htm
It's stuff you can just go buy in a shop, we've used it here for around 15 years to connect across a street to the other office. We have a laser interlink.
Now I have another thing to implement for Bring Your Own Device...
This does make me wonder, however, if we could see fiber optic gbics that don't cost thousands of dollars each if the technology that makes this free-air communication possible can be adapted to fiber optic applications.
Do not look into laser with remaining eye.
Welcome to Slashdot
How does the Mbit/mW compare to a 802.11b/g/n pringles cantenna?
Which can achieve further distance assuming LOS?
They demonstrated one way data transmission over a very short distance, not a network.
"Eve of Destruction", it's not just for old hippies anymore...
Laser based FSO isn't exactly a new field.
1Gbps data rate with a diode laser isn't that hard to achieve even with pretty simple drivers and 1-bit amplitude modulation.
Neither is using wavelength multiplexing some revolutionary new idea.
So... huh?
How much is that in sharks per second?
if your short range fiber gbics cost thousands of dollars each, you're buying from the wrong vendor.
Try $50
imagine 4 bits per cycle, then imagine 16 or 32 or some other power of 2. Then we can apply a sort of Frequency Division multiplexing with diffrent colored lights, so for a single strand of fiber times that by 4.
Last year there was an article about thumb sized atomic clocks with the ACCURACY to potentially make this feasaible.
Petabyte+ class single cable link anyone?
Oh, and I remember how famously obnoxious IrDa was to use. a few feet? sheeeeeet, the devices had to be virtually touching and you needed software that supported the link. Might as well use a null modem, and cary the cable. Before the age of USB sticks, readily available residential grade ethernet products, etc, but AFTER the age where the 1.44mb floppy disk was relatively useful, transfering files between computers was a royal pain in the ass to begin with.
Ok, the real question is... how does this apply to /.'s new BI focus? Can I use this instead of spreadsheets or specialized software to properly align my Business Intelligence with the synergies of the corporation for maximization of profitability?
Ouch, that hurt...
"The urge to save humanity is almost always a false front for the urge to rule." --H.L. Mencken
This was done years ago. I remember seeing the story, I think it was on gbppr. The problem is, these laser pointers aren't designed to be used constantly and they wear out.
"A red and green laser pointer were used, each transmitting a stream of data at 500Mbps, which is then multiplexed at the receiver for a grand total of 1Gbps"
That made me think of blue lasers, which would have even better rates.
But, how about longer waves, such as infrared or even radio? Are there any radio lasers around? THAT would do for long distance calls, and proabably be enough for E.T. to phone home.
I had assumed an LX range with mode conditioning, and on top of that I had assumed that something that could work through a medium as imperfect as ambient air with ambient light and still achieve speeds of half a gigabit could achieve much faster speeds over the controlled conditions of cable, like say, 10gb over laser-optimized OM3... Which are currently thousands of dollars.
Do not look into laser with remaining eye.
It's only a matter of time before the MPAA/RIAA gets this outlawed because pirates could be using it to broadcast entire ripped DVDs to each other in mere seconds using sharks with frickin' multiplexin' red and green lasers attached to their heads! You laugh, but it will happen.
A proof of concept on laser pointer networking was done two years ago, if you are interested see
http://www.diva-portal.org/smash/record.jsf?searchId=4&pid=diva2:325270 - Fulltext at
http://www.diva-portal.org/smash/get/diva2:325270/FULLTEXT01
is getting the sharks to hold still.
I'm surprised the cell phone companies haven't implemented something similar on their towers to reduce backhaul. Have dozens of towers in a given area relay optically to a super node tower with amazing backhaul. Have them relay to a few others in a standard mesh network layout for redundancy. Might even reduce their spectrum need if they are using channels to talk tower to tower. May have some issues with rain I suppose though, but that could be mitigated if laser wavelengths for which water is not refractive exist. Or just use laser arrays with heavy multiplexing and parallel signal reinforcement.
For a start, visible (and invisible) light has a frequency of between 400 and 800THz (800 and 375nm), which is unlicensed spectrum worldwide.
My God! They're broadcasting my movies over an unlicensed, unregulated carrier! This MUST be stopped! This "visible" light will aid paedophiles, piracy, terrorists, drug dealers and all manner of criminality!
Time flies like an arrow. Fruit flies like a banana.
not good for 20% of males that have red/green defective colour vision, you desensitised clod...
There was an unknown error in the submission.
They already use laser diodes in these things, so using a laser diode won't bring the price down. It's really nothing new here. Just somebody has decided to play with lasers.
http://www.laser2000.de/index.php?id=370844&L=1
This 600m kit is typical of short range systems, notice the prices? 300 euros.
http://www.laser2000.de/index.php?id=374687&L=1
The bigger ones, e.g. 1.5km is only 900 euros
http://www.laser2000.de/index.php?id=370040&L=1
this technology: http://news.slashdot.org/story/12/05/02/1744223/finally-a-shark-with-a-laser-attached-to-its-head
10Mbit, 1200-1400 meter range, GFDL-licensed open designs: https://en.wikipedia.org/wiki/RONJA
Instead of lasers they use LEDs with relatively inexpensive lenses.
So it shouldn't be too difficult to just use one laser for each side of a full-duplex link at 10-100 Mbits and just send standard Ethernet frames back and forth. Not for any serous applications, just to prove it would work.
I don't think ambient air with ambient light is a particularly imperfect medium at those speeds. Nothing in the ambient blinks that fast.
apk, you're a tard
A group of students at The University of Pretoria in South Africa did exactly this while I was still studying there, this was circa 2001.
A large part of their motivation was to help build a technology for high-speed networks that were not subject to the state protected telecoms monopoly.
They used almost exactly the same technology, lazer-pointers for sending streams, but I believe they used solar-cells for receivers.
I remember they boasted speeds of over 1mbs which (back then) was incredibly fast (in fact faster than the internal buffers of the P2 computers they used - so that the data actually slowed DOWN after being received) but I don't believe they ever went beyond a single point-to-point connection.
Maybe one of the students who were involved is on slashdot and can give more details ?
Unicode killed the ASCII-art *
I'd like to see a beowulf cluster of these! Wouldn't you?
The paper describes phase measurements. Phase of what? They don't use any modulation scheme that involves phase -- they have dumb on/off "modulation", their carrier is light itself, and their detector is a photodiode, it can't measure phase of light without an interferometer that they don't have.
As light passes through the air, random (but low-frequency) "noise" modulates its amplitude and introduces varying delay. If we are really unlucky (fog, dust, very long distances), there will be multiple paths, "smearing" impulse response. However light itself has frequency in Terahertz, single-bit signal with all its harmonics is up to Gigahertz, so this is nowhere close to what happens in RF data transmission. This kind of distortion just works like a simple low-pass filter, limiting the data rate. No filtering or amplification can affect that, so we have a data rate hard limit for this kind of "modulation" there, and it's much higher than what they are trying to do.
On top of this, receiver is very much nonlinear -- all those harmonics are introduced not by medium (air is almost perfectly linear as far as light transmission is concerned) but by the receiver photodiode. THAT can be compensated by "adaptive filter" that at the same time tracks shift in amplitude -- levels slowly drift (clouds of dust and fog may appear and disappear), but no one really cares how linear or nonlinear anything is, because "amplitude modulation" was by unpredictable noise anyway, and all we need is to discern levels for 0 and 1. Now, look at the eye diagram in the article. It's clear as day that one on the left is smeared because of amplitude drift (so traces start at slightly different points) and one on the right has "filter" adjusting to the changing amplitude (so all traces start at the same point relative to the "true" moment of level transition). It's level adjustment, plain and simple.
Oh, but what about phase?? There is no "phase". Phase of the carrier is gone after the photodiode. There is no frequency to apply "phase" to, pulses' frequency is arbitrarily chosen by the user, and nothing at all depends on that frequency as long as it doesn't reach Terahertz range. What happens is simple delay, that also may drift when changing air pressure changes refraction coefficient or, to put it plain, speed of light in the air, or when transmitter and receiver move, or when un-compensated amplitude changes shift the time when waveform (electric, not light) crosses given levels. The only "phase adjustment" is synchronization with those pulses' leading and trailing edges -- if this is "adaptive filter", then RS-232 is also "adaptive filter" because receiver constantly adjusts its time when it receives start pulses.
So here it is, the transmitter does nothing but turn laser on and off forming pulses. The receiver receives the signal with amplitude and delay drifting, mutilates it further by introducing its own nonlinearity and immediately loses everything related to phase (and frequency) of the carrier because carrier is just light falling on a photodiode. Everything past that point is just level adjustment (what they call adaptive filtering for amplitude), Schmitt trigger (that paper omitted, so maybe they forgot to put it there) and synchronization (what they call adaptive filtering for phase).
Oh, and both their drawings are wrong. And their amplifier doesn't do anything for "compensation of nonlinearity" that simple level adjustment wouldn't do better because -- surpise -- it already contains an amplifier with gain adjustment, and all we care about is two levels.
I don't agree with the original poster about the deficiencies in IrDA.
This was in a time when dial-up access was the norm, if you were lucky you had ISDN, and your cellphone gave you patch 9600 baud connectivity. IrDA was fine in this situation.
I used my Nokia 6320 phone and Palm V to restart servers whilst on call from the comfort of restaurants, and even to make changes to Perl scripts from a different country. The range was poor, but fine, the performance was limited by the cellphone not IrDA.
Kids of today, they've no idea
You may think me a tired, old, cynic. I'd have to disagree about the tired bit.
Now the sharks with friggin' laser beams attached to their heads will be able to enjoy, not just a warm meal, but also high-speed digital electronic data communications!
Sorry but this has existed as a communications system in the middle east and india for well over a decade now. People over there have been doing this with laser pointers and LED's for a very long time.
Granted it was only 100bt as it used existing transciver chips, but the jump to 1Gps is not that hard.
and why red and green? The existing designs all use RED for alignment and then IR for data comms so that it's not visible from acute angles.
Do not look at laser with remaining good eye.
Fully opensource 100mbps version right here http://ronja.twibright.com/
Rather old too.
The ham radio experiments have modulated the light wave with a low frequency radio signal - say 25kHz. This would imply a more linear output stage than a pulsed light output stage - or a modulation technique such as FM which doesn't care so much.
Maybe this is the way we'd use QAM - in the modulated signal. I wonder if the maths applies as it does for normal radio waves - if the spectrum of the output will be the normal AM or FM spectrum so even a very pure light source once modulated will contain other colours. Thinking of the maths I'd expect so, though it's light and colour so it would seem counter-intuitive and I wonder what other effects may come into play.
Worked great for me. For extended range or in 'noisy' environments id stick an optical fiber between the 2 devices.
---- Booth was a patriot ----
Those with a (or even an) historical bent may be interested in the first outdoor optical communication system, the heliograph, which used reflected sunlight for long-distance communication via Morse code. The record distance covered was 183 miles (295 km), between Mount Ellen, Utah, and Uncompahgre Peak, Colorado on 17 September 1894.
To my knowledge, this record for terrestrial (i.e., non-moonbounce) optical communication has never been broken, even by modern laser and LED systems. The closest attempt of which I am aware is 179 miles (288 km), between Mount Horror, Tasmania and Mount Liptrap in South Gippsland, Victoria, Australia on 27 October 2009, using non-line-of-sight techniques (they bounced Luxeon LED light off of high cirrus clouds, and used very-weak-signal digital coding and modulation techniques).
Do not look into router with remaining eye.
Oh wait...I get it now - they put lasers on sharks so that they can have a network across the ocean! (http://news.slashdot.org/story/12/05/02/1744223/finally-a-shark-with-a-laser-attached-to-its-head)
... move around? Or do you have to stay in the same spot where the lasers were aligned?
now we need to go OSS in diesel cars
My experience, on windows: enable IrDA on phone. Put phone next to laptop. Phone recognized, systray icon pops up. Send files.
IrDA was the closest we ever came to solving the still-unsolved problem of how to transfer files wirelessly between two machines sitting next to each other. It's telling that the de-facto standard now is to carry around USB flash drives: god help you if you've lost whatever cereal-box prize you were using.
Compared to the dicking around we have to do with bluetooth - which, incidentally, is still brutally slow - IrDA was awesome. It wasn't fast, but files weren't big ten years ago.
employing red and green laser pointer lasers (LPLs).
Well that's fucking clever of them.
Man, after ten years of watching the internet slide into being something resembling a privately-run prison system, a light glimmers.
Damn it, yes. Use da lasers. We need alternative cheapnets that go around the massively centralized and monitored leylines that the commercial and government spooks have utterly co-opted.
Lasers on the roofs, encrypted comm, backbones sneakily flashed around the countryside. Move the switches and bridges into software on commodity hardware, so they can't be controlled by manufacturers working with spooks. Move DNS into decentralized networks. Build an alternative to the US-controlled internet. Goodness defined.
Illegal in 3... 2... 1... but don't let that stop you all. Sex outside of marriage was felonious once. So is betting on cards and ladies showing their ankles. What is illegal today is a joke tomorrow.
Now, pity we can't use the former TV channels.
Me and a buddy of mine cobbled together a Super Sekret Spi Laser Telephone System in highschool. It was a couple of cheap laser pointers and photovoltaic cells wired to the transistors in radio kits with microphones. They only worked well at night, over shortish distances, and were a complete bitch to aim. Still.. SUPER SEKRET SPI LASER PHONE!
You mean a glimmer in the REMAINING eye.
Hardware is just a small part of the cost. There is installation and maintenance. The cost of going more than 25M includes installing a pole ever 25m, getting power to the poles and maintaining the repeater on top of the poles. Even solar panels need to be cleaned regularly. Would you rather maintain 2 switches inside buildings and 100M of cable or 16 laser nodes( for redundancy) of which 12 are on top of poles with attached solar panels? All it takes to bring this network down is for a birds with long tails to perch in in the nodes.They may not even work in high fog conditions. LOS lasers too fragile for important networks.
Currently our internet connections are very bursty. Most of our network seems to have been built with the assumption that most people will not use all of the pipe. Maybe the ISPs can split traffic into two portions: fixed rate guaranteed wired bandwidth, and a much, much higher "you're lucky its sunny out" blazing fast variable bandwidth.
I went to the paper, expecting to find a fairly high powered laser that is not a pointer, and expecting to call someone out on calling them pointers. However, they're only 5 mW, which is indeed a pointer. Cool that they can use such low powered lasers for this.
The FDA has regulations in the U.S. saying that no laser products over 5 mW may be marketed as "pointers":
Class IIIb lasers cannot legally be promoted as laser pointers or demonstration laser products.
I don't know why this is being hyped so much... from my brief look it seems pretty dodgy.
I'm not an expert in data transmission, but I have reviewed quite a few papers.
Two main points stand out:
1. They have two lasers of different wavelengths just so they can use the phrase "wavelength division multiplexing", but the lasers point at separate photodiodes! The lasers could be the same wavelength and it would make no difference.
Doing this adds nothing to their paper and lowers my impression of the research quality.
2. Their adaptive filter seems to require that the receivers already know the correct data in order to measure the amplitude/phase error and adapt.
Why would you need to transmit the data if it is already known at the receiver???
I would reject this paper.