Hackaday Explains: Li-Fi & Visible Light Communications

A new way to transmit data is coming that could radically change the way that devices talk to each other: LiFi. Short for Light Fidelity, LiFi uses visible light to send data, creating the link between router and device with invisible pulses of light. This type of Visible Light Communication (VLC) uses something that is present in pretty much every room: an LED lightbulb.

What is LiFi?

Li-Fi sounds like the an engineer’s fevered dream: it is fast, cheap, secure and simple to implement. Speeds of up to 10Gbps have been demonstrated in the lab, and products are now available that offer 10Mbps speed. It is cheap because it can use a modified LED lightbulb. It is secure because it only works where the light is visible: step out of the room and the signal is lost. It is simple to implement because it uses an existing technology: LEDs.

The basis of the technology is in turning the LED light on and off very fast. By switching an LED on and off millions of times a second, you can create a data signal that can be detected by a sensor, but which is invisible to the human eye. At the other end, another LED detects these pulses, and can send light pulses back in response, creating a bi-directional link. If you combine this with wired Ethernet or a WiFi network, you have an awesome combination: an Internet connection that uses visible light for the last link.

One way to think of this is like an Infra-Red remote control, but on steroids. Your IR remote works by turning an IR LED on and off at a frequency of about 38KHz, encoding data at a speed of up to 120bps. A sensor and a microcontroller in your TV detects this light, and converts it into commands. Some computers and PDAs in the 1990s used a faster standard called IrDA, which could send data at kilobit speeds. IrDA never really caught on, though, although the group behind it is working on new versions that could up the speed to 10Gbps.

Li-Fi uses a similar idea to IrDA, but with visible light instead of IR. This has the advantage of being a wider chunk of the electromagnetic spectrum, and with some clever encoding, the red, green and blue elements of modern white LEDs can be used to increase the amount of data that can be sent, a technique called Color Shift Keying (CSK). Because the data is being sent by LED light sources, it can also be focused. Literally: you can use lenses to direct the signal.

The proponents of this technology are planning to build networks using small LED devices on the ceiling of a room which they call attocells. Because they use visible light, these attocells can be shaped to transmit and receive over a very small (and focussed) angle, so they won’t interfere with each other like WiFi devices can. In a recent paper, the proponents of this technology claim that, in a typical 400 square meter office, this could deliver between 12 and 48Gbps to devices: significantly more than a WiFi or other wireless network that has to deal with interference from other similar devices.

Another interesting new development is combining a Li-Fi signal with a solar cell. With a bit of clever engineering, the solar cell can both power a small device and receive the Li-Fi signal. So, a device could receive both power and data from the same source, which could be very useful for Internet of Things devices or small sensors.

Are There Any Standards for Li-Fi?

At the moment, there are no real standards for Li-Fi. The systems that are currently available are partly based on the IEEE 802.15.7 standard that defines the physical layers of the network, but this is rather out of date. Standards are coming, though: the Li-Fi consortium is the main group working to create a WiFi-like standard that will provide the same sort of interoperability that WiFi devices have. There is no timeline for this at the moment, though.

How Can I Try Li-Fi?

There are two ways at the moment: buy one of the first devices, or make your own. PureLiFi was the first company to sell a LiFi device: the Li1st. Since then, the company has also added the LiFlame, a first version of their attocell concept. Using a device on the ceiling and a USB transceiver, it can send and receive data at up to 10Gbps.

The other alternative is to make your own. Disney Research has published a paper that shows how they built a Li-Fi like device into an existing light socket using an Atheros SoC board running Linux. Unfortunately, they have not published any schematics or code for this.

Hackaday reader [jpiat] has, though: his project uses an Arduino to control of an RGB LED light to send data using a modified 1-Watt LED light. His system only works in one direction, but it is very simple: the Arduino uses simple encoding on the data, and three data lines to control the LED. It is robust enough to send at a decent speed, though: he claims that it can manage an impressive 1kbps speed using a 1-Watt LED light to a distance of 3 meters. That’s an impressive glimpse of our Li-Fi future.

The home-made VLC system built by [jpiat]
The home-made VLC system built by [jpiat]

61 thoughts on “Hackaday Explains: Li-Fi & Visible Light Communications

  1. This is neat, but for internet access this technology will never be embraced by consumers. The advantages over traditional wifi disappear when you consider that you’d have to install one of these bulbs in every location where you want to get a wireless signal. That would make the cost and frustration factor too high for anyone who wasn’t really excited about technology. Also, your lamp has to get the internet from somewhere. The only practical application I see would be spreading tiny sensors around your kitchen that got a signal from a light fixture on the ceiling, which could actually be really cool.

    1. Yep. There are even more basic limitations that will annoy consumers. You can’t use it in the dark and you need unobstructed line of sight between your device and the LED bulb.

      1. You can use it in he dark but you’d need a different transmitter and receiver…two light bulbs and two different filtered receivers. Also I hate doing this but in the top paragraph it says that it’s invisible, but you’re using visible light…just doesn’t make sense. Also, aren’t radiowaves invisible. I dont understand how you use the invisibility thing as a ploy to get people on board with this. I don’t really see how this will really take off except in niche markets. I can also see this as being a huge security issue as it is probably easier to build a visible light receiver and place it somewhere to do a purpose of infiltration then it would be to do it with a wifi connection, also visible light typically travels further than a wifi signal. So if someone did want to have a secure network from people outside, that means no windows. Which may not be that big of a problem for some companies that already do that but I’d bet that companies that deal with mostly business administration don’t and therefore would make prime targets for malicious hackers.

        1. Well, if you have good encryption, it’s really not much less secure than Wi-Fi. Many businesses will have their buildings wired for Ethernet anyway, so they probably don’t use much Wi-Fi (or Li-Fi) for critical tasks, and it’s not hard to corral that part of the network so it can’t interact with critical systems.

          1. Well you can have a sneak peak at a window on a building and get the LiFi signal easily from far away. Compared to wifi, where you would need directional antenas and some serious gain.

    2. I think the issue is more the devices that use it, it’s not easy to keep a laptop in line of sight of a lamp but for things like tablets and phones it’s even worse. Or you need to use a superpowerful light (wide beam laser) on it, but at a wavelength that you can’t see, so IR, but pumping lots of power in IR drains the battery you would think.
      Bu if you just use bounced light won’t you get interference?

      You know desktop computers used to have connectors for IRDA ,modules long ago, it was discontinued because nobody used it really, because it’s light and needs line of sight. And now we pretend it’s all new.

      And IRDA went up to 1Gbit/s, so the same damn speed as this LiFi? Uh…

  2. Um, I remember infrared ports on laptops sharing files and also digital cameras that could upload to them via the infrared link. A lot of motherboard used to have a set of headers on them reserved for this. So we up the transfer speed and rename it “li-fi”?

  3. I have a biology question related to LiFi. Is there a difference between invisible to the eye, vs not perceived or processed by the brain? In other words, is there potential for other side effects of having bright LEDs flickering at these rates, outside of our perceived visual spectrum?

    1. For that sake, I share a bit of personal knowledge: every single light bulb out there nowadays is driven by a switch mode control circuit, what means it already flickrs (at a rate of anything between 10kHz and 500kHz I imagine). I am no biologist or phisician, but I can say that we are already exposed (with led light bulbs) to any effect that li-fi might have when it is available.

    2. Just try to think. Any “old” neon bulb flickers at 50Hz or 60Hz. And from some people it is also perceived, provoking headaches in few cases. Almost all modern devices (smartphone, monitors, car lights) are enlighted with LEDs modulated in PWM (flickering, even at low frequencies like 100Hz).

      The flickering higher than 100Hz is filtered out from the retina, and what you see is actually what you receive in the brain.
      Exception is when you have to adapt under a sensible (at first times, at least) flickering, provoking headaches, i.e. old CRT monitors or failing “old” neon bulbs. Another exception is when the intensity of light is not well adjusted, or the color is not well proportioned to the ambient.

      In fact, the LiFi it is just a modulation of the light: assume that WiFi is dangerous (I’m not saying that it is), then is dangerous “a priori” rather than depending from the modulation used to transfer bits, i.e. the type of data.

      Despite this, studies demonstrates that the wi-fi modulation due to transferred bits, in their turn used to download data from hackaday.com pages, are strangely revitalizing on mind and body.

  4. Am I the only one who keeps my phone in my pocket? Li-Fi seems great for direct line-of-sight connections, BUT occlusion is a real serious issue.

    For that matter, at those kinds of speeds, are reflections off of walls an issue? Think multi-mode vs single-mode fiber-optic cables.

  5. So, it’s like a fiber-optic link, only without fiber. I guess technology has some some cool applications, but it is too much different from Wifi to replace it entirely. Most people have their WiFi equipment in the hall and computers in the rooms, LiFi fails here immediately. Unless you want to strech LED strips around your house. Which kind of nullifies “wireless” part of the deal.

    1. If I was the guy in charge of this product I would design some kind of device that would bridge between the internet and the house mains. Then, anyplace you want Li-Fi, you screw in a Li-Fi enabled bulb. The bulb just connects to the internet through the socket its screwed into. You would need a filter on the mains so the devices in your house didn’t interfere with your neighbors, and vice versa. Might want to silver tint all your windows too, or invest in curtains….

      1. Having re-skimmed the paper, it would seem to not be a hoax.

        “4.3.1 Results of the Survey of Devices. Dial-up and leased-line modems
        were found to faithfully broadcast data transmitted and received by the device”

        And yes, encryption is a vital part of protecting the data. Potentially still some metadata, etc. I was merely challenging the assumption that the light (data) would not leak outside the room.

      1. People always ended up paying more rushing out to pay money to be beta testers. Sometimes a technology fails and the companies drop the support and sometime even better one from a competitor comes along. “Future proofing” and “investments” do not work for tech toys as there are alway better and cheaper ones that comes along. Yesterday’s top end product will be a middle or low end product. Look at what your needs are and make sure it is good until the next upgrade cycle.
        As for broadband speed, until there is a fundamental regulatory and market change where I live, the regular WiFi routers would be more than enough for the “high speed” internet thank to the duopoly.

  6. Well, there is some qualitative difference between IrDA (or infrared remote devices) and LiFi, as infrared uses amplitude, and LiFi frequency modulation (or CSK, whatever). I don’t see why infrared couldn’t use some kind of CSK, but it’s not a matter. I am much more interested in will LiFi be immune to pulsed LED lighting noise.

    1. Precisely. AFAIK that phosphor also has some persistence, it continues to glow briefly after the light is turned off, which would tend to degrade signals. I wouldn’t expect a solar cell to be great for receiving high frequency signals either.

      And if you want your light bulb to continue to function as a light bulb, then you can only use a small fraction of its power to modulate data. Otherwise when on, it would dim when LiFi is in use. Or when off, it would visibly light up. Where is the advantage of sharing a 10W LED for both light and data, if you’re ultimately limited to using only, for example, 100mW of that for data? A separate 100mW LED would cost pennies, and might even simplify the design by separating power and signal paths.

      Not to mention, 2.4Ghz WiFi has pretty good bandwidth for normal tasks. Your average IR LED is what, around 300Thz? So shouldn’t that, even if not pushed to limits, provide several orders of magnitude more bandwidth than WiFi? Why the push to go to visible light?

      I just don’t get all the hype about LiFi.

  7. I have seen this bullshit in so many places. it works, but it is not practical. What is a good way to bring that 10GBps link to my lightbulb? as long as it requires another cable to bring the internet to the bulb, it is dead.

    Their product is in no way better than placing an AP on the ceiling in the room, having one in every room.

  8. Thanks for publishing my hack ! The 1kbps was achieved using only the red channel of the 1W LED. This 1kbps does not transfer to 1kbps of useful data as the the data frame contains synchronization symbols and bytes are encoded on 10bits (added start stop to detected sync loss). The protocol needs to be improved as it only allows transfer of text data (for now) and does not support addressing. Some users were able to reproduce my results with some kinks in the way that helped improving the code a lot. Other users replaced the receiving LED with a proper light sensor for improved reliability and data-rate. Next step is to buy a 10W bulb and a modified receiving LED to avoid emitter/receiver alignment problems.

  9. I cannot, for the life of me, see how PureLifi can market this as a replacement for wifi.

    If you want Internet access you will need to have a connection from the room you are in to the room with the router/cable modem (which means Powerline Ethernet, Wifi, or an Ethernet connection) — that means LiFi is not and cannot be a replacement for current technologies.

    Not to mention the fact that the article mentions:
    LiFlame, a first version of their attocell concept. Using a device on the ceiling and a
    USB transceiver, it can send and receive data at up to 10Gbps.

    So you want to take a 10Gbps connection and bottle-neck it, both ways, to 480Mbps or less ??

    I can only see Lifi catering to a very small nice market (at least in the way they are marketing it now).

  10. Data on modulated light is neat. And [Jpiat]’s example project is very cool.

    But when you want to get up to many megabits, or cover longer distances, things get challenging. The one hacker example of a good LiFi-esque high-speed connection that I know of is the (now very old) Ronja project: http://ronja.twibright.com/ These guys actually got 10Mbs over a kilometer fairly reliably, which is no mean feat.

    What I learned from looking into this eons ago: lasers aren’t as robust to atmospheric conditions as a big, wide LED beam, sensors and optics really matter when you’re shooting over long distances, and the modulation/demodulation stuff is a lot like radio.

    LiFi (or whatever) should be a lot easier b/c it’s indoors over short distances. Making it portable and practical without large power requirements or a very focused beam is probably non-trivial. Will this ever be better than WiFi except for a very-low-bitrate service? Dunno.

    1. I am curious what exact phenomenon you are referring to with “lasers aren’t as robust to atmospheric conditions as a big, wide LED beam” ?

      Can you elaborate the advantage of LED over LASER as the light source in this application?

      The only thing I can think of is gaussian beam waist ( https://en.wikipedia.org/wiki/Gaussian_beam#Beam_waist ) vs divergence ( https://en.wikipedia.org/wiki/Gaussian_beam#Beam_divergence ) : the smaller the waist the larger the divergence. But if this is the argument (as indicated by “big wide LED beam”) then it is a poor one, as choice of light source is orthogonal to choice of beam forming optics… For example, to measure the distance to the retroreflectors on the moon laser light was used, but the beam was first expanded to have a large and far away waist to decrease divergence and hence increase intensity at the reflector.

      If you understand another important reason why the LED as a light source is preferable over a LASER, I am curious to understand such other reasons!

  11. But there is SOOOO much noise in the visible spectrum…. that’s going to be a problem. Unlike the RF domain, EVERYTHING is a reflector, waveguide, absorber, and/or emitter of light(even the theoretical “black body” – even the air)… true, not in the entirety of the spectrum, but enough of it to matter. Hell, even ambient physical temperature makes a difference more so than in RF(especially in the IR)…

    This is why high-bandwidth photic signaling is done through optical fibers, where things are controlled and predictable(more so than atmospherics, at any rate)…

  12. The last widely used version of infrared communications was FIR, Fast InfraRed, at 4 megabit. The various speeds are listed here. https://en.wikipedia.org/wiki/Infrared_Data_Association Looks like they aren’t giving up on the technology, a protocol supporting up to 10 gigabit is in the works.

    What knocked FIR off laptops and second generation smartphones (early Palm OS and Windows CE phones had it) was Bluetooth. Longer range and no line of sight limitations.

    A high speed IR system’s advantage over a visible light system is it’ll work in the dark, without lighting up the room with data.

  13. Now we need this implemented in the backlight of a TV to send data back and forth to rear surround speakers and smart remotes. Having a remote that can tell you what channel it’s on and still last months on a battery would be great.

  14. If this tech doesn’t gain popularity, it might be used by companies for ultra secure intranet structures, best way to keep data safe is to use a media (now a connection method) few use, or even know about. That’s why I still have a few reliable 5.25 drives and disk laying around. Neat.

  15. im torn because im obsessed with LEDs…

    on the one hand it’s neat and a SWEET hack that i plan on trying (with huge 10W RGB)
    on the other hand, it sounds like an investor scam.

    WHAT IF THE INTERNET USER ALSO HAS A COMPUTER SCREEN LIGHTING UP THE ROOM??? does the 10gb transfer drop down to 10mb or 10kb because the guy turns on the very laptop(screen) that would be using the data ??? what about during the day with the window open? would it work?
    is this only suitable for a torrent download? … only goes reasonable speed once screen is off? what if somebody is using a normal tv? normal ccfl or led ???

    im not trying to *%^%) on any hack, this stuff is INTERESTING!
    but whatever happened to the idea of converting (IR) fiber to visible fiber?
    if you have 10 wavelengths of light you get 10 times the bandwidth, no?

  16. The technology is in its developmental state but soon will be the time when to access internet what we will need is to switch on the light bulb. The speeds are astonishing to hear but will we actually get those speeds and is there any future for lifi.

  17. I can share some of my research that can give you all insight in multi-channel VLC model that I develop and that will be presented on some future academic conferences :)
    Any way if you are Anxious to know how all of that work in real world please wait I will pull all my research to my hackaday.io profile later this year :)
    “Analysis of Visible Light Communication System for Implementation in Sensor Networks”

    p.s. 1-2Mbps is some thing that all can get with out much… :)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s