A fast and cheap optical version of Wi-Fi is coming

This is a lot like infrared-based networking. Practically, you're just substituting visible light for IR -- and IR has the advantage of working in areas you want to be dark.

One can view a whole cityscape busy street but not see the pickpocket working the crowds.

One can smell a lie even though there is no odor.

Women living together do not smell anything peculiar, but get olefactory clues that synchronize their menstruation.

There is UV and IR wavelength light that we do not see, yet we sense from heat and sunburns.

Imperceptible does not mean senseless. We view a lot of things; but we lack critical awareness of details.

That's no different than today's Wi-Fi devices where anyone with a radio can listen in on your Wi-Fi communications.

The solution to thwart Li-Fi eavesdroppers is the same solution used to thwart Wi-Fi eavesdroppers: encryption.

Most Wi-Fi communication today uses standard AES encryption. So, you can listen in to my Wi-Fi communication as much as you want. But, unless you have the password to decipher the encrypted communications, it will all look like a random jumble to you.

Fundamentally this is nothing new, TV remotes have been using this for years, although at very low datarate, and even today you can still find IrDA devices, they just can't compete with the range and datarate offered by RF based links.
Maybe fast IrDA will make a comeback, maybe not, I would still put my money in UWB for now, but anything is possible. Eitherway, if TE is worth anyone to pay to read, it should at least do the proper research and point out how this is different from IrDA, rather than trying to portray an old idea as the next great invention.

This is a very old technology that does a few things well (such as your TV remote control).

Although IBM have tried hard 20 years ago the data rate used in WiFi mode (that is with a range of a few metres) is low by WiFi standards, even though the radiated power is relatively high (WiFi base stations only emit 100mW). People are keen to try because it 'feels' right (people are also looking at 60GHz for the same reason) but in truth even 5GHz WiFi is not so great. It could be different if signals were beamed' but this is tricky.

But fear not - there is really plenty of RF spectrum around top do GHz comms - we just need more of it assigned to WiFi ('unlicenced') - write to your MP/congressperson.

Sweet idea. Now we just have to kill all the world's moths to ensure perfect data transmission.

Flashing lights can rarely induce seizures.
Or worst yet cause common head splitting migraines.
And it may not even be perceptible light flashes, but some sub-perceptible effect on the retina.

Is a female coworker in the next cubicle tangling with PMS and frequent migraine complaints worth having the next gen Li-Fi?

I wish I too could dare make a comment without knowing the first thing about the science. Or do I?

I don't have much to say to the article itself, but couldn't leave this allusion to Terry Pratchett's novel uncommented :)

For everyone who wants to see before they can believe it, here is Prof. Harald Haas at TED demonstrating how the technology works:

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

It stated "imperceptible by the human eye." You won't have those problems.

Reminds me of synchronising and transferring files between my early laptops and Pocket PCs (Smartphones without the phone bit) through infra-red light.

The major challenge is always (Bandwidth)x(Distance). Bandwidth of an LED is limited by the capacitance of the device, while output depends on the size of the LED, increasing capacitance. The problem is made significantly more difficult than fiber transmission because the light is not spreading out in three, not one, dimensions.

"...light bulbs—some 14 billion of them around the world—are almost everywhere and often on. As they are gradually replaced by LEDs..."
1.Almost everywhere...gradually replaced.
2.If not now, when?

This is a wondeful idea, but it's predicated on the assumption that LEDs will be adopted world-wide almost kimmediately.

SHowever...ome of Mr. Edison's original light bulbs are still plugged in. We love our light bulbs. We're not gouing to give them up in the near future.
FACIT LUX...LENTE

Data transmission with laser has been available, albeit very expensive for a while..but it is cumbersome. This is really cool for short range transmission and also broadcasting: in malls, offices...one can see several apps

This is an incredible technology that will take off big time -- invest in companies that are pioneering this!

Imagine the potential data transfer rates... this could radically reinvent the way people interact with the internet.

I'm excited about this technology!

I think it may be the perfect broadcast channel to beam commercials to nearby pedestrians and audiences.
It is the next best thing to putting commercials in your dreams.
Makes you want to never sleep.

It could be useful to have a tablet wirelessly connected to a powerful computer, so the computer does heavy processing tasks, and send back the answers via LED lights. That way a consumer could carry a tablet to any room on his house or office, meanwhile accessing far more computing power than available on his tablet.

To an extent, this would likely be true, but then again, I don't think it would be a significant increase. My wifi router is on 24/7, as are most others'. We just don't turn them off at night. But it's easier to flip a switch and turn off your light. Let's say your entire house's lighting system is wired into the network (probably not as expensive to do as it sounds). You really only need one light in each room turned on, which you can power off when you're finished using it (something most energy-conscious people already do).
A physical modem would still need to exist, of course, which would maintain any hard-wired systems. However, most people already keep lights on most of the time they're using their wireless device. And the difference in electricity savings between incandescent and LEDs may completely offset this anyway.

Of course, those like me who prefer natural sunlight to indoor lighting during the daytime may still need to utilize some sort of hybrid system, like mentioned above, between LiFi and Wifi.

I see this application as a very viable one at least commercially. Businesses keep their lights on constantly during operating hours, and the extra power requirement to run a lifi network would probably not be significantly more (possibly less) than that of a wifi router.

Of course, the flaw with airplanes is that usage might be restricted mostly to daylight hours. At night - especially on long flights - the cabin lights are off most of the flight, and I'm sure there would be many complaints from those attempting to sleep if everyone using the plane's LAN had to turn their chair light on!

Nice article. But I will raise some point -
1. Can LED replace SON-T in street lighting now if we consider the Lumen efficiency of SON-T?
2. Can LED replace CDM-T Lamps with CCT more than 90?
3. Can LED replace incandescent Halogen Lamps (12V, x-K, x-Degree, x-Watt) with CRI = 100?