
mercedes are trialling LED headlights
which feature many (70?) active steered LED's would be quite trivial to modulate at a hundred megabits per second, Car2car photon based comms look set to be around 5 years away then!
Forget about Wi-Fi - the future of home wireless networking is, according to boffins, the light bulb. So say a number of researchers and technologists who are looking to light to provide the next step in high-speed data networking in the home. Reg Hardware Mobile Broadband Week The principle is simple: turn a light on and …
Audi and Citroen (at least) already have LED running lights, and most manufacturers seem to use LED for brakes these days too.
If car-to-car comms were only used to say "Sorry, my bad, didn't see you" and "Thanks!" then I think it'd be a good idea. Not sure it would be limited to such polite messages though....
Car-to-car comms could be used to find out traffic conditions ahead of you; average speed, density of vehicles, traffic jams, roadworks, roads closed (signs could put out their own signal) etc. by cars automatically transmitting data to each other.
But a significant amount of cars would need to have the system installed for it to be any use.
Indeed. So many car drivers claimed they didn't see their motorcycling victims that the TRRL designed 'leg protectors' for bikes. However, it became clear that these devices would cause the rider more damage than the alternative. Having experienced this 'didn't see you' excuse and the 'blind spot' claim (there is not a blind spot, merely stiff or lazy necks that should not be behind wheels) I chimed when I saw your response.
As to 'Li-Fi' itself, a data thief's paradise, and it is not good saying that encryption will solve the problem. Already we have WEP in the morgue with WPA in cold pursuit, and I have a feeling that WPA2 will not be far behind. I network through the plugs and it is a delight to experience.
While the idea of using LED's broadcasting around the room (IR or visible) should be fine, the idea of using the white LEDs lighting the room isn't. The bottom line is that while the driver LED itself can be modulated at high speed, the photoluminescent phosphors generally cannot. The commonly used photoluminescent phosphors have characteristic exponential decay times of microseconds to milliseconds, depending on the exact phosphor. The most common phosphor in white LEDs, YAG:Ce3+, does have a pretty short persistence time of around 0.1microseconds. So maybe one could get data rates as high as a megabit per second, but that would be about it. But remember, the efficiency of white LED bulbs depends crucially on the phosphors chosen - so choosing ones for data transmission could easily negate any energy efficiency advantages of white LEDs over, say, CFL's.
Been here before a few times haven't we?
I seem to remember Apricot/Sirius used to offer optical networking back in the 80s, then IRDA came along late 90s/early 00s and promised to take off.
Can't wait, this time we've got the electro-sensitive bunch who will discover a brand new bandwagon of lumino-sensitivity to jump on and sue from.
it's Infrared- as used in millions of devices for decades? But with visible light? And slower? And your lights need to be on to use your LiFi device at home? If you'll excuse the pun in my sarcasm, that sounds Brilliant. There's nothing I like more than having some irritating light at the side of the room letting me use my new Wireless keyboard, or playing computer games in a well-lit room so my phone can sync.
Which retard thought this up, and how long will it take before they come up with the "revolutionary" idea that they could use an invisible wavelength of light for it?
One of the massive advantages that visible or IR light comms have over WiFi or Bluetooth- or even NFC- is security. You can have broadcast wireless data comms all over a building- and keep the contents secret by closing the blinds. Want to make an IR pay-wave system secure? You get coatings that would scatter any 'leaking' IR and stop it from being intercepted.
Plus it's almost certainly lower-power than Bluetooth for 'normal' speeds. And at Gigabit speeds is almost uncontested in the wireless space.
And using an IR one would mean you can control your TV/SkyBox/etc without needing the daft WiFi/ Bluetooth combination my Samsung supposedly-smart-TV requires.
tl;dr: IRDA FTW, LiFi Sux, Rant over for now.
Quite a good rant: I'm not sure what your point is.
Data via light is no more inherently secure than any other electromagnetic transmission. What you suggest is the normal "security by obfuscation" approach which is known to be crap. Want security then add security.
The efficiency of electromagnetic transmissions is also largely independent of the frequency. Power is directly related to data rates and distance. Want efficiency then make it directional. Given the known heat losses still inherent to LED I doubt very much that it is more efficient than Bluetooth.
Point one:
Security through obfuscation is NOT known to be crap, at least not in this instance.
To explain: With Bluetooth or WiFi your neighbours can see your traffic (assuming you're in a regular urban/suburban environment). They can't necessarily read it- you'll have encryption turned on- but they can see it. And breaking encryption is just a matter of time/processing power anyway. You can even outsource it to the Cloud these days.
With an IR network, your neighbours aren't even aware of your network- or at least can't see your data. Those above/below you have inches of concrete / wood / plaster and carpet between your network and them, and those across from you can be blocked by closing windows / shutters. I'm sure there's a range of glass coatings out there to block out IR as well, so for most 'home' applications you could even have the curtains open.
On a smaller scale, say touch-payments, it would be enough to have a squishy O-ring seal around the photodetector, or have it in a box (same idea as some Chip-And-Pin pads that have a 'mask' over the top).
Compare this with NFC cards which can be read through a pocket- or RFIDs that can be read from many feet away.
No interception of your wireless signal means there is NO external access or interception. That is a proper physically-secure wireless network, and on top of that you can add further security.
Point two:
Bluetooth (class 2) broadcasts over a sphere up to 10m in diameter. IRDA has a beam width of 30 degrees (less than 1% of the covered volume and a 'standard' range of 1m (though according to Wiki the newer standards include longer ranges).
Also, Bluetooth seems to drain power even when not being used. IR just sends data when you want to send it and receives it when you want to receive it.
So yes, it'll be lower power.
Point three:
If you have 3 IR LEDs in a known arrangement any mobile phone nowadays should be able to give a full 3D position fix just off the camera. So could be deployed to provide a low-cost 'local GPS' system for use in emergencies.
Or, more likely (and more profitably), you could have an Augmented Reality display like a dancing leprechaun in front of an irish pub. All while beaming the menu and current drinks promotions to you wirelessly- without interfering with the broadcasts of the shop next door.
Yes, this requires a bit more infrastructure in the real world, but then so does owning a Website or adopting NFC or printing up QR codes.
Point Four:
Nearly every gadget in my house (and, amazingly, car) has an IR remote control. With IRDA back in smartphones you could control your various devices just using the IR connectivity rather than using bluetooth/wifi as modern TVs etc seem to want you to.
Point Five:
IRDA is nowhere near good enough for a broadcast infrastructure like TV or mobile phones and I'd not claim it is. Even for wireless headphones it's not as good as BT.
But as an addition to a mobile phone, it's better than a lot of the currently "in" things. And even helps with advertising, providing location-specific location, etc.
Point six:
IR is far superior to LiFi, because, amongst other things, it doesn't need your lights to be on to use your Internet at home.
Rant re-stated!
I'll take your word for it....
BTW how do the 3 IR LEDs give you a low cost local GPS???
Assuming you are talking about triangulation and line of sight, I would need the phone to have either a 360 degree panoramic lens (if I'm in the centre of the triangle) or a pretty accurate gyro + level to figure out the changes in angle made while swivelling around the camera to look at all the 3 lights.
@Smileeee,
There is a technique for using a telescope to see the scan of a CRT screen in an office, it was posted on the Register many years ago I believe. Rumour had it that the CIA was using it. With the rise in LCDs the technique probably became redundant.
On the topic, I think this is a fantastic technology, it will allow for at least simplex communications to augment WiFi.
And what if I want to wander into the garden with my laptop because it's a nice day? No network.. heck, even go to the bog and it drops out.
There's a reason IRDA failed - it requires line of sight and that is a huge disadvantage of such technology.
The flashing the mobile screen thing to send a message probably has legs - but as a general networking solution it's a nonstarter.
I'm not sure IRDA failed as a whole since it's still used in so many TVs but I guess you meant in terms of data transfer.
Taking the TV example a step further my PS3 remote is bluetooth and there is a definate lag on first use. I'm not sure if that lag outweights the benefit of not needing line of sight since I'm always sat in front of the TV when using the remote (Play TV).
I still wanting to point it at the TV anyway... havn't trained myself out of that yet !
it was slow as molasses.
Worked with some guys who had tried this (data network over LED) in the early 90s. The worked at it for a good while but never managed to get it working. reliably or at speed. It's that pesky box in the diagram labeled "Signal Processing"...too much multipath, too many nanoseconds and picofarads to get a decent data rate. Not to mention all the optical interference.
This is essentially what I thought..
While I don't currently have LED lighting, as I don't think it would be very environment friendly to replace every energy efficient bulb (all CFL atm) with LED bulbs, I wonder at the practicality of this.
The reason is that most homes don't have lights on during the day, and even a lot of companies are encouraging people to switch off lights if they aren't needed. Some companies are even going as far as to use light pipe systems to distribute the light from the sun, or one bulb to several rooms.
I'm sure there's more to it than the article says, but why would this be better than HomePlug/Powerline? It looks like the lamp is being modulated using the mains, but then you have a gulf of space filled with thousands of other light sources or reflectors before you get to the sensor, which then has to be in an optimal position to balance the signal-noise ratio.
Sounds less mobile than Wi-Fi and less resilient in a non-mobile solution than HP.
Unless I've missed some very obvious piece of genius.
It would need to be modulation of low voltages really so the technology for the lighting modulation wouldn't need to affect the mains. However how do you get the data to the lamp in the first place? Perhaps you might use a power line technology, or if you are doing a new installation then you might be able to use copper or fibre.
How come the article avoids mentioning the obvious flaws? (Or the solutions if I'm wrong.)
* As a few commenters have already said - you need the lights on.
* Strong sunlight will probably swamp the signal.
* What sort of light source will the return channel need? A tiny LED will probably not get though. Do you need a USB-connected desklamp aimed at back your MiFi router?
why use the electromagnetic spectrum when we can do it with high frequency sound ?
okay the data rate might not be great.
On the plus side it will also annoy the hell out of any animal that has a hearing range
greater than humans... lots of psychiatric dogs and cats running around because they're being sonically tortured. yippee !
So in few words, you want dial-up, except high pitch?
Isn't dial-up about 20Bits a second or something similar. I don't see it taking off.
Light is good, its the fastest thing we can "control" in our universe and therefore ideal for data transfer. But in a cable makes sense because optic fibre can carry the light and there are no issues with line of sight. But wireless, all you'd need to do to corrupt a file is throw a mirror in the way.
Face palm at the Li-Fi concept. And a pair of leg warmers for you, my commentard, for that blast from the 80's!
IIRC the last modem I had was able to do around 42k/s even the earliest audiocouples could handle 150 bits/second. (I have to wonder if you where really around in the 80s if you think is was "20bps or similar")
The limiting factor in the case of the modem is the limitations on your carrer. the phone system was not designed to handle frequencies above a certain threshold, and therefore there are limits on what a modem is ALLOWED to do to prevent too agregious electro-magnetic interfearence.
sonic transmission would not generate these types of interferance, and therefore should be able to (legally) push beyond the limits of my old modem.
"Light is good, its the fastest thing we can "control" in our universe and therefore ideal for data transfer. But in a cable makes sense because optic fibre can carry the light and there are no issues with line of sight. But wireless, all you'd need to do to corrupt a file is throw a mirror in the way."
Excuse me, you're standing in my network.
Also, how are you going to power the 'bulb' when you cut power to the socket?
There seems to be some misconception about 'the lamp needs to be shining bright' for this to work. This is not, in fact, the case - it's well possible to dim the light down very low indeed and still get data across. In fact, data transmission should be easier in a dark room - less interference. But even in full sunlight, a dim lamp will be able to transmit data without it being visibly 'on'.
The question isn't the data rate you can cram into the visible spectrum - it's hundreds of THz wide - but avoiding interference from all the other bulbs, and also feeding data to the sockets.
This would make some sort of sense for additional downlink bandwidth for media streaming. Except, as many commenters have pointed out, most people turn the lights off to watch films. If you're putting a light-based modem into lightbulbs then why not stick some IR LEDs into them? You don't need the visible light LEDs switched on and you get essentially the same coverage and bandwidth.
I'm sure this is probably covered by others, but .. seriously - this sounds great!
So there I am, putting on my media PC in the lounge room to stream a video onto the TV off some form of networked device (a fileserver in a distant room/youtube/some internet source/whatever). Now, when I watch movies, I like to turn off the lights for that full 'cinema experience'.
Nope, can't see a single problem with visible light design :/
I'm sure there will be tests, I'm sure most of us can predict the results...
But all lights flicker, and just because some scientist using a brand new bulb in a lab can't see it doesn't mean I can't see the flicker from the much older tube in the office. (Though no one else seems to be bothered I need to counter it with sunlight to stop the colourful hallucinations induced by sitting under it 10 hours a day)
Late to the party, but I've seen the tech first hand. The flicker is a non issue - I am quite sensitive to flickering lights and don't notice anything when looking at them working, and the light is not modulated on-off, it's a much more complicated modulation scheme than that.
Also the light can still transmit data while it is not perceptibly on, the data rate might be lower but it can cope with quite high noise to signal ratios.
The people worried about moving from room to room: I think the lights will operate somewhat like mobile phone towers, on a more localised scale, where your connection should automatically switch to the most visible light source. This is just speculation on my part however.
Disclaimer, I'm loosely related to the Edinburgh project.
IIRC Mitsubishi have patents for flashing the keyboard backlights of phones for reception by the ambient light sensor on other phones, again low bandwidth data like your phone book entry.
Likewise fluorescent lights in shops and meidal facilities have been looked at as LF comms for mobile devices, patient tracking and updating price labels in near real time.
The key phrase seems to be "smart ballast".
This sound like a gimmick to sell more LED's but I think non domestic lighting has a) a *huge* installed base and b) is fairly efficient. That makes a *lot* of market inertia.
Personally I quite liked a system to allow a daylight patrol of street lights. Any council vehicle could have a van top box that at normal road speed could pick up a data packet of up to about 1500 bits. More than enough to report if the the various parts are failing/failed and schedule a call out. Yes in many places the only way the local authority knows if a street light has blown is if it's reported by the public.
"This sound like a gimmick to sell more LED's but I think non domestic lighting has a) a *huge* installed base and b) is fairly efficient. That makes a *lot* of market inertia."
Two possibilities:
1: "Riding the wave". The guys behind this are running on the assumption that once LED pricing hits a certain sweet-spot, there will be a massive changeover, possibly pushed by new legislation on energy efficiency and/or mercury in manufacture and/or the workplace. If they have a mature(ish) technology before that happens, then they have an "in" to the market. They need to have their technology available for when everyone rewires, cos they're not going to rewire twice.
2: "Making the wave". They may feel that the increased efficiency of LEDs isn't going to be enough to convince people to swallow the cost of switching, but that this technology can help push the uptake of LEDs. "Double your office wireless capacity while halving your electricity bill." That's a lot easier to sell....