So... It's a glorified IR connection, then? That's so last millennium.
Everything old is new again.
Scientists in China are reporting great success in replacing traditional Wi-Fi radio traffic with Li-Fi, a system that uses the light signals from an LED lightbulb to provide a line-of-sight data stream. Li-Fi, or light fidelity, was developed by Professor Harald Haas at the University of Edinburgh, who was disenchanted with …
I must say I like this better from a security point of view. Once outside the room your pretty much cut off from the network, of that light source, Unless connected via another point. outside the room. However its currently much to slow for practical application of byod. Not to mention the unanswered questions that sit around interference and physical device location. (considering my tablet Lives in a case this could be a problem)
Will be interesting to see how this matures as it makes its way to the market.
Twenty years or so ago, I worked with some folks who had tried something similar in the office environment.. They were plagued with coverage and multipath problems, as well as the low signal-to-noise ration and the large size of the required detectors. Uplink wasn't an issue, as the transmitters weren't in the lights, but tacked on top of cubicle walls, using the ceiling times as passive reflectors.
It never worked to spec. I don't think networking-by-light has ever worked in free space, at least I have never seen a successful multipoint network with anything but carefully aligned lasers on rooftops.
// light meter in one pocket and telescope in t'other
"In the future you will not only have 14 billion (LED) lightbulbs, you may also have 14 billion Li-Fi's deployed worldwide for a cleaner, greener, and even brighter future," Hass said at the TED talk introducing the technology."
Firstly, LED lightbulbs are by far most commonly of a single led construction, with a phosphor that changes the colour to whatever shade of white you choose. The phosphor will prevent any useful on/off communication because it's just too slow. RGB leds have the flexibility to change colour, and they're fast enough for this on/off comms, but they cost a lot more. Saying that you "might" also have these LiFi's in their place is bullshit - no-one's going to pay for that.
Secondly, I want to second the other poster's quote about being a glorified IR connection - that's so last year. Heck, even RS232 is still in use today.
Third, and the real clincher, if you have the word "Green" in your marketing blurb, you automatically tell me it's a complete pile of crap. Mainly because you're so desparate to sell the shit, you've resorted to words that bring an emotive response without actually meaning anything.
I don't know what you've been smoking but it's clearly hazed your vision or the only led lights you've seen have "cool white" or some such and not clear covers. I just came from class where they had replaced all the florescent tubes with led tubes last year and I can tell you for certain that the led tubes have over a hundred leds each. Anyone can go ahead and verify that fact with a simple Google image search but don't let reality get in the way of a perfectly good uneducated rant.
If you want to bitch about the color spectrum of led lights, fine, but whinging about the use of "Green" in the marketing is just as childish as the folks who whinge about something not stating that it is "Green". Let's face it, too many people have been taught the buzzwords without really knowing the meaning so saying something has high "whatever" efficiency means absolutely bupkis to the more than 90% of people who don't really understand how what they are buying works but if you compare it to the hulk and say "it's green", at least they think they understand. Not saying it's wrong to call bullshit on marketing that won't back up its words with data but to not understand that most of it is hyperbole for the great unwashed smug fuckers is, IMNSHO, just silly.
Alas Eddy you're the one who doesn't know what they're talking about.
When John Tserkezis said "single LED construction" he's clearly flying over your head; this doesn't mean "lights only contain one LED"; he means the light is made using a single-COLOR LED die and is internally using a phosphor coating to convert that single color into to white light. There is no such thing as a white LED die, as LEDs are by nature monochromatic. This is how all 'non-color-changing' white LEDs work.
His point (which is correct, and a very good one) is that the phosphor used for this conversion has a vastly slower on/off response time than the LED itself, and this screws up fast data comms.
Color-changing LED bulbs use Red+Green+Blue (and sometimes more) dies and modulate them to provide different colors. However the 'white' output (imo) is typically unattractive, and they're more expensive and complicated to make than a regular phosphor-conversion white LED.
Ah, thanks for the quick lesson. The reason I still take classes is to keep from falling further behind even as I approach retirement but unfortunately as science and technology is ever expanding and my memory seemingly ever contracting it's not going to be easy.
I remember HP or someone like that had an IR office network system something like 20 YEARS ago. they had ceiling transponders that looked much like camera domes, and plug in transonders for a variety of computers.
this system was, of course, something slow (I'm remembering 1-2Mbit/sec shared bandwidth), and yes, I think it used IR. line of sight transmission created various problems.
As pointed out before the phosphors of "white" LEDs are a problem.
Then if you go to higher modulation speeds, reflections become a serious problem.
So anyhow you need seriously specialized technology to get this working well enough so it'll be able to compete with WIFI. It's not just "modulate the LED light you already have".
Well there are still some differences between what we can easily do with radio and what we can easily do with light.
The bandwidths with Wi-fi are small enough so we can just digitize the whole signal and perform digital calculations on it. That way you can, for example, do an FFT on it to use a modulation scheme called OFDM... which helps you solve the problem of reflections.
Even if you use fairly narrow bandwidths for light, you'd still need highly precise carriers you can modulate your signal on. With radio it's fairly simple to build a finely tunable 2.4 GHz source which will have most of it's energy within a band only a few hertz wide. With light a single colour LED is essentially a noise source covering several terahertz of spectrum. That's like the old spark-gap transmitters. You can transmit something, but any serious sort of modulation is pretty much out of the question. Essentially with LEDs and even simple laser-based systems everything is at those standards.
Today we can actually use more advanced modulation schemes, but that's far from your "LED-bulb". You need highly precise lasers and extremely fast data processing. This works great for fibreoptic lines where the only thing you need to compensate for is the difference in frequency between the transmitting and downmixing laser. Compensating for reflections is probably out of the question for yet a few more years.
So we are getting to a point where we could do Wi-Fi with light, but it's not going to be doable with LED lights.
"So anyhow you need seriously specialized technology to get this working well enough so it'll be able to compete with WIFI. It's not just "modulate the LED light you already have"."
Yeah. IRDA was the thing for palm top wireless connections to printers.
Small transducers, stopped by walls, no licensing issues. It had it's attractions.
Relatively slow by modern standards...
Modulating the light isn't too much of a problem, assuming that an ultraviolet filter can be used to minimise signal degradation through phosphor glow. The main limitation with this sort of approach is, or at least a few years ago at least was that cheap and cheerful photo-detectors only get a frequency response up to MHz speeds if you're lucky. Detecting high frequency signals at low light levels tends to get rather expensive.
Detecting high frequency signals at low light levels tends to get rather expensive.
My guess is that the signal will have to be modulated somehow. Much like TV remote control IR and IRDA used to. This would actually be manditory to isolate the pulsed light noise you're likely to encounter in this type of environment. Heck, I've seen TV remotes become entirely ineffective due to badly designed compact flourecent bulbs.
Getting 150Mbs under lab test conditions is entirely plausible, but under real-life conditions? The biggest problem is going to be the difference between light on, and light off, at the receiver, under ambient (sunlight through the window) light conditions. It comes down to signal to noise ratio. TV remote IR is not very tolerant here, but they generally don't have to, they just tell you to shut the damn shades and be done with it. But if this turns out to be like IRDA was, where the two devices have to placed "just so" to make it work, people are going to get annoyed - we've been spoiled with highly functioning devices nowadays and most of us will probably gravitate to a technology that "just works" rather than what's sexy. (apple products excluded - that's another subject)
Using your car headlamps, no one will be allowed into the city without the required iDlight's fitted and your road tax status can be checked from half a mile away at ground level or from satellite using the angled reflectors mandatory for all new vehicles.
This is obviously to reduce the incidence of terrorists driving into town and normal law abiding citizens have nothing to fear.
Given the line of sight limitations, I would have to think this would best be used in two ways: broadcast data (which might be better served with some kind of broadcast radio data band) and point-to-point connections where wires and radio are unsuitable. It could have a use in security applications where a controlled wire-free link across an air gap might be needed for temporary transmission of data. Depending on the receiver sensitivity, it might also be a cheap alternative to laser links that have been used between skyscrapers.
I wonder if the tech could be used as a successor to IRDA, capable of transmitting and receiving more information at a time than Bluetooth and barcodes while still in a confined setting.
Controlled settings are fine.
How about in office cubicles where you don't want to fiddle with physical connections and 150Mb/s to each desktop is ok. It saves sharing 802.11ac with the whole floor and having to deal with the vagaries of radio signal propagation. If you want to go Apple-esk, you could do a magnetic connector to ensure no leakage but still have no mechanical clips to snap off.
Likewise in a house you can do 150Mb/s per room with fewer problems from overlapping signals.
But in both scenarios, you'd need to place the two ends of the link in ways that may not be so practical.
An office would be in a better position to use physical connections because most have access to a drop ceiling which alleviates the hardest part of the wiring process (a conduit pole can get the wires from ceiling to cubicle). Since a cubicle link would have to be put on the ceiling anyway, it would probably be easier (and perhaps more secure) to wire up.
As for the home, layouts can be more random, making the system less practical than a WiFi. Range is becoming less of an issue with more powerful access points.
It's quite possible to ask for it in new homes and major renovations. I mainly wired my downstairs when the kitchen was being remodeled, as they tore the old inner wall down, allowing me to work around the studs (getting from the attic all the way to the exterior conduit would otherwise have been impossible due to twists, turns, and staples). It's when you have to deal with "in situ" situations that wires get tripped up.
I think I've worked out a solution to the inherent security issues surrounding transmitting light into the office space. Put a bit of fibre channel chable between the two LEDs... capture all the light from each LED so that they only way any light can escape is down the cable... hey presto no more light leaked data. Now all I need to do is figure out how to stop some twat clipping an transducer onto the cable and capturing light from protons which refuse to go around corners.
Biting the hand that feeds IT © 1998–2019