News 100x Faster Than Wi-Fi: Li-Fi, Light-Based Networking, Standard Released

[jasonf2]

The line of site issue makes this a pretty limited use case. I could see it as a backbone link standard to connect buildings or point to point stationary links, but I am pretty skeptical on any implementation in a mobile device that the phone case( or pockets) won't make this pretty much useless. That neglects the fact that you will also need a stationary beacon installed line of site which gets expensive really quickly.

 

[PEJUman]

Yes, which is why they'd put enough sensors to cover all directions.

If early VR headsets could use "light houses" for tracking, which also required continuous visibility, then it seems plausible optical communication can work for it. You would probably need at least two transmitters, in case one of them got occluded by your arm or something.


If there's so much smoke or fog in your room that it's hard to see through, you've got other problems.


When visibility is blocked, your phone could just fall-back to wi fi or 5g.


Isn't IrDA point-to-point? I think this is a cocktail party protocol, allowing for one-to-many or many-to-many.

Do you own a wigig device? I do. it's not about smoke/fog in the house. it's about how sensitive the bandwidth & latency to interference. 228 Gbit/s @ ideal, vacuum chamber, 1~`99 gig @ nominal condition would suck for anyone trying to develop towards its high bandwidth/low latency capabilities.

Visible/near visible light is problematic because so much of our life already uses this, which means there are a ton of near visible interference around us that this thing have to build ECC for.

 

[bit_user]

Visible/near visible light is problematic because so much of our life already uses this, which means there are a ton of near visible interference around us that this thing have to build ECC for.

Just like any RF technology, you would need good error resiliency.

However, the frequency band is the main thing that saves it from all the interference sources you mentioned. A high-pass filter works wonders for that. In fact, the sophistication required of the filter is probably a lot lower than what's commonly used for RF communication.

 

[razor512]

Companies have been trying this tech for many years now, and it never makes it to market. The issues they always run into is output power restrictions, and interruptions when whenever anything crosses the beam.

I with they would have continued developing 802.11ad WiFi instead, but tune it for even higher speeds at even shorter ranges. One benefit at the time was effectively 2.5Gbe like performance over WiFi at a time when other standards were struggling to do 800mbps real world throughput (basically since many routers at the time that offered 802.11ad had only 1GbE ports, you need to download content simultaneously from 3 wired clients, but you could get well over 2Gbps without much issue.

Imagine if they continued developing the 60GHz band with more modern enhancements such as higher QAM modes and other improvements. They could probably get a 10+gigabit connection for devices within around 10ft of the AP, which could be useful for laptops, e.g., imagine you have a device set up where once in close range to the 802.11ad AP, it switches over and then begins backing up the data on the laptop. Or utilize it in a smartphone where when charging and close to the 802.11ad AP, it quickly backs up all data on the device to the user's NAS.
I currently do that now with foldersync pro. though the process would be so much faster with even faster WiFi.

 

[velocityg4]

light is just visible RF.... recall 60 GHz nearfield wifi? Intel AD/HTC vive wiress devices anyone? aka WiGig.
very high bandwidth, can barely pass paper/carboard obstacles?

Visible wavelength range is 400 ~ 770K Ghz, aka 400~770 Thz. Talk about line of sight & stuff better be transparent in-between the devices (like smoke/suspended particles/humidity/etc).

Seems like the answer is to use gamma rays. Unbelievable high speed transmissions and pass through all sorts of objects. With only a minor drawback to your DNA.

 

[SKFB]

It also gets into the question of why bother. You do not really need super fast speeds to a device like a cell phone or pad device you will just fill the storage quicker. Its not like you are downloading microsoft flight simulator to a phone.

If it is a desktop computer and it must be in the same room as the router you might as well run a ethernet cable and avoid the complexity.

Only thing I could see is maybe vr units so you could have only the display in the headset to cut the weight.

Try selling the idea of using a LAN connection with a laptop to a consumer who considers everything computer related as the equivalent of literal magic. You'll quickly discover that they will fail at best and absolutely violate that port at worst. So this technology is also highly useful for offices and consumers using portable devices, as they can get a secure, fast and reliable connection, without using a cable (and without making IT people lose their sanity)

 

[SKFB]

Companies have been trying this tech for many years now, and it never makes it to market. The issues they always run into is output power restrictions, and interruptions when whenever anything crosses the beam.

I with they would have continued developing 802.11ad WiFi instead, but tune it for even higher speeds at even shorter ranges. One benefit at the time was effectively 2.5Gbe like performance over WiFi at a time when other standards were struggling to do 800mbps real world throughput (basically since many routers at the time that offered 802.11ad had only 1GbE ports, you need to download content simultaneously from 3 wired clients, but you could get well over 2Gbps without much issue.

Imagine if they continued developing the 60GHz band with more modern enhancements such as higher QAM modes and other improvements. They could probably get a 10+gigabit connection for devices within around 10ft of the AP, which could be useful for laptops, e.g., imagine you have a device set up where once in close range to the 802.11ad AP, it switches over and then begins backing up the data on the laptop. Or utilize it in a smartphone where when charging and close to the 802.11ad AP, it quickly backs up all data on the device to the user's NAS.
I currently do that now with foldersync pro. though the process would be so much faster with even faster WiFi.

Does your TV remote stop working when someone stands in front of it? Cause your TV remote uses IR to control the TV

 

[SKFB]

This!

I can see it being very good in an Office setting and even a subway tunnel where it's hard to get wifi access. This will definitely be more commercial than an end user at home.

Same. I think if anyone are to use this at home, it's gonna be IT peeps and DIY'ers who then specifically rig the setups to always get connected. I wouldn't be surprised to see a HAT, pHAT, Shim or similar with LiFi capabilites in the future, as it can be pretty good for either further automation or better control of desktop CNC machines, regardless of whether they're additive or subtractive. Being able to control these machines from elsewhere, without using extra cables, as well as getting increased speed, would be absolutely amazing imo. It would allow the user to control CNC machines in their sheds, from the comfort of their home office PC, simply by piggybacking off of the existing electrical grid

 

[SKFB]

needing to use a flashing light source and reciever sounds like wireless optical fibre.. but without the fibre you need to have a strong light flashing constantly into an area and the client end need to flash back to communicate, I can't really percieve this to be useful outside of say, commercial storage server for hot swarpping drives. for civilian use having a overhead light keep shining and my phone/computer shine back don't seem like a good idea for eye fatigue or light pollution

Infrared light isn't visible though...

 

[bit_user]

Seems like the answer is to use gamma rays. Unbelievable high speed transmissions and pass through all sorts of objects. With only a minor drawback to your DNA.

Aren't they also known for corrupting your data by flipping bits in RAM (and presumably also NAND)?

I'd expect they're highly directional, as well. So, you'd either need to spray a lot of them, in all directions, or have some way of aiming them.

 

[lcw]

Has anyone actually tried to read the standard in question or at least its summary? 802.11bb is up to 9.6 Gbit/s. Slightly lower than 224 Gigabytes/s, isn't it?
It is slower than Wi-Fi 7 which you can buy today.

from https://lifi.co/lifi-speed/

"Lastly, the kind of light can also affect LiFi speed. Data rates as high as 3.5 gigabytes (GB) per second using a single blue LED, or 1.7 gigabytes (GB) per second with white light, have been demonstrated by researchers. This new technology can, in theory, offer speeds of up to 224 GB per second."

They claim "the more light you put in a room, the faster your connection will be" which suggests the expected use of wave division multiplexing. In fact, their quoted 3.5 GB/sec rate multiplied by 64 shades of blue = their theoretical 224 GB/sec.

Everything at lifi.co mentions visible light, not infrared. They claim "its [sic] invisible to the human eye due to the speeds of how the light flickers" but they also claim it can operate as low as 60 lux which is about the level of typical house lighting. (I guess that means security-conscious environments can install a manual "kill switch" near the entrance to each room)

The IEEE's 802.11bb standard originally looked at 380-5000nm, which includes visible light through microwves. The approved standard seems to stick to the infrared 800-1000nm range. That's probably a wise choice to gain adoption, but it does mean lower theoretical bandwidth.

 

[razor512]

Does your TV remote stop working when someone stands in front of it? Cause your TV remote uses IR to control the TV

A TV remote uses a pretty wide angle IR light output, and runs at an extremely low data rate.

Compare it to other old standards that used IR LEDs, such as IrDA, as data rates increased for the standard, the range dropped and they became more sensitive to obstructions and corruption.

On some old laptops with them, they could get up to about 4Mbps over IrDA, but they had to be very close, without much light contamination in the spectrum being used, and they were sensitive to anything that would partially obstruct them. The LEDs on them if viewed with a camera, would be much brighter than a TV remote.

 

[bit_user]

On some old laptops with them, they could get up to about 4Mbps over IrDA, but they had to be very close, without much light contamination in the spectrum being used,

About 30 years ago, my HP graphing calculator had an IR link capability. I think it was still faster to use a serial cable, so I hacked one together. I might be imagining this, but I seem to recall some idea where a teacher could use IR to broadcast multiple students' calculators.

IIRC, they ran a 4-bit microprocessor at 2 MHz and had a base RAM of just 32k. And yet, they still managed to pack an interpreter for a LISP-like language and there were a couple Gameboy-caliber games you could find for them.

You can now get an Android app that looks & feels like the old HP 48G.

 

[plm999]

LiFi might be useful for driving Lightfield display panels. HDMI is currently constrained to 80 GB/s, which means a fair bit of decompression hardware is still required in the display to render. Imagine a remote rendering box, e.g. loaded with many TFLOPS of GPU goodness to render the Lightfield image, then using LiFi to beam directly to the panel (or panel array) or even a headset. Now the panel can be thin and light and sit flat and pretty on the wall. Or lightfield goggles that can be lightweight and run on battery for hours instead of an onerous umbilical. The big kilowatt sucking heater doing all the work can be across the room (or at the other end of an optical fiber in another room).

The other use case is high security environments. Glass can easily block the IR, so it doesn't leave the room. This use case would be much less bandwidth centric, so reflected light might still reach all of the laptops in a conference room even if there's a minor obstruction. Powerline and such would not be helpful, as that would be relatively easy to steal. If neighbors stealing bandwidth is an issue, you'd probably have to pipe optical fiber from your router to each room, which might be a bit impractical, when you can just select a more secure password instead.

 

[bit_user]

LiFi might be useful for driving Lightfield display panels. HDMI is currently constrained to 80 GB/s, which means a fair bit of decompression hardware is still required in the display to render.

Current tech just uses a microlens array fused to an 8k panel. It would be cost-prohibitive to get out ahead of the conventional display panel industry.

 

[Volker Jungnickel]

Dear commenters,
unfortunately, the author mixed two things: What LiFi can theoretically achieve (224 Gb/s) and what the 11bb standard defines (9.6 Gb/s). 11bb sends the same waveform defined for 11n/ac/ax over light, using an electrical subcarrier signal at low frequencies. 11bb can use MIMO and WDM to increase data rates and robustness in the optical domain, same as 2.4/5/6 GHz can do this in the RF domain. As it is based on existing Wi-Fi specs, 11bb can only reach the same data rates 11n/ac/ax can reach. With 11bb, Wi-Fi can use the light as an additional frequency band, but will not change its operation principles. Light is more secure and reliable than RF, and thus a nice complement to RF in some applications like in industry, aircrafts or a hospital.
Kind regards - Volker