Infrared may no longer be a punchline, as IEEE approves 9.6Gbps wireless light

https://www.theverge.com/23795599/li-fi-wi-fi-802-11-bb-internet-of-light

by el_hacker • 3 years ago

141 147 3 years ago

theverge.com

hilbert428• 3 years ago

Why such a puny speed?

dist-epoch2• 3 years ago

It's faster than most computers wired Ethernet.

coltonweaver2• 3 years ago

The theoretical limit of WiFi 6E is 2,500Mbps: https://www.highspeedinternet.com/resources/wi-fi-6-vs-wifi-...

This would be nearly 4X that.

hilbert428• 3 years ago

Perhaps so, but this new standard reminds me of the born-dead USB1 standard, which at the time of its launch was so slow as to be almost useless.

This new optical standard is only about 50% faster than the now-ancient copper SATA standard of 6Gbps. It has stuff-all reserve for future development!

Every time we have one of these new standards, Wi-Fi whatever, they are just a bit above the previous standard and never at the limits of the tech. There is never enough reserve to give them decent longevity. It's as if those manufacturers involved in setting standards were making a standard that would be quickly obsolescent—that is, they'd guarantee ongoing production/sales without much effort.

It's been such a consistent problem for so long that it has shades of the Phoebus cartel about it. As always, the user ends up paying more because of premature (planned) obsolescence.

Pet_Ant• 3 years ago

Well the speed is the speed of light, ideally this should have low latency. The datarate will be effected by how quickly you can turn on and off the light and that will have some limits.

bragr• 3 years ago

As apposed to radio waves which also travel at the speed of light?

manmal• 3 years ago

Latency in WiFi comes from interference and contention (wait until other devices have finished sending; this has improved with Wifi6 AFAIK), and some other, probably less significant factors. Travel time of the signal through air should be the same.

sp332• 3 years ago

The LiFi page claims much higher speeds. https://lifi.co/lifi-speed/

hilbert428• 3 years ago

Right, now look at this standard by comparison.

zokier3• 3 years ago

That 224GB/s figure is some serious bullshitting. For comparison, the fastest optical ethernet transfer over fiber is 800Gbps or 100GB/s, and even that is over 8 fiber pairs and still in-development tech. Claiming to do over twice of that in free-space with "common household LED light bulbs" is pretty ridiculous.

fulafel2• 3 years ago

People seem to be conditioned to the choppy and slow wifi speeds, so slower-than-Ethernet speeds can still sound appealing in comparison.

hilbert428• 3 years ago

Right, I think conditioning is definitely part of it, but my point was (as I hinted elsewwhere) the move to optical would have allowed for a vastly better performance with little additional overhead.

coltonweaver2• 3 years ago

Am I correct that this would require direct line-of-sight for any device using an antenna[0] enabling this? It seems a bit impractical for your typical consumer use-cases if so, but maybe beneficial for some more commercial scenarios.

Would love to know if I am wrong here though! It seems very interesting.

0: https://www.purelifi.com/products/light-antenna-one/

dist-epoch2• 3 years ago

The idea is to put it in the ceiling (lightbulbs).

Light bounce also works, so if you are in a typical room the walls also become antennas.

hfkwer• 3 years ago

So if you turn your back on the ceiling light you lose your connection?

zadler• 3 years ago

Years later, we’ll find out that it messes with people’s sleep…

Havoc2• 3 years ago

Or pets can see it and don’t appreciate the disco

BenjiWiebe• 3 years ago

Modulating it at several hundred Mhz, the pets won't be able to see it flicker.

gnicholas• 3 years ago

Would this be easier to intercept from a distance?

letitbeirie2• 3 years ago

Probably harder in a practical sense - its signal doesn't travel through walls or most commercial windows so the intercepting sensor would have to be located in the same room (if indoors) or have a heroic signal/noise ratio to avoid getting washed out by sunlight (if not).

gweinberg• 3 years ago

Strange title. If infrared is the punchline, what's the joke?

antonvs2• 3 years ago

I imagine the idea is that infrared communication itself is a joke, if you need higher speeds or have any objects in the path of the beam.

PartiallyTyped2• 3 years ago

Does anyone even need 9.6Gbps wireless?

supertrope3• 3 years ago

Wireless laptop dock is one application. A 1080p monitor signal is 3.2 Gbps. Dual monitors is 6.4 Gbps. One "4K" display is 12.5 Gbps.

Wireless is notorious for quoting numbers that include protocol overhead and assume a near zero noise and congestion environment. The typical layer 7 throughput of 802.11 is about half of the link rate. Of course it goes down from there without line of sight in the same room.

sedatk• 3 years ago

640Mbps ought to be enough for everyone.

dvngnt_• 3 years ago

wirelessly streaming vr

PartiallyTyped2• 3 years ago

It seems that VR is the only reasonable application.

readams2• 3 years ago

I'm curious who has a problem that this solves. You need to have line of sight to something very nearby. So you'd need to wire 10Gb Ethernet to the LiFi points.

But then why not just use wired Ethernet?

ugjka• 3 years ago

airports would be one case

NortySpock• 3 years ago

Conference halls, open plan offices, LAN parties, maybe data scientist teams sharing large datasets around a room...

afavour• 3 years ago

But then why not just use wired Ethernet?

This is a recurring question and one often answered by “people rent houses/apartments and can’t just run cables through the walls”.

I’m not all that sure what I’d use this specific technology for over Wifi but I can understand why people don’t use wired Ethernet.

mabbo2• 3 years ago

I own my condo but have this exact issue. There are 1Gb Cat5 cables inside the cement walls and I cannot replace them if they ever break, nor upgrade them.

Already this is a problem as I have 1.Gbit internet.

Edit: I am silly and wrote Mbit instead of Gbit.

Nullabillity• 3 years ago

How do you plan to provide line of sight through those same walls?

Cloudef• 3 years ago

I have apt with only 1 ethernet plug, but coax run to every room, so i used moca adapter to have wired internet everywhere, works great

supertrope3• 3 years ago

The same reason people use Bluetooth audio. It's more convenient because there's no cord. (But there's a battery to manage, interference, pairing, and inferior audio quality).

antonvs2• 3 years ago

Just a vaguely related anecdote, I consulted for a company that had two offices on a river, in old converted houses, separated by a marina. Running a cable proved problematic for various reasons, both bureaucratic and relating to the nature of the site.

To connect the networks between the two buildings, they had a laser connection set up in the widow's walk on the buildings. It worked great except when it was foggy or raining very heavily, or when the marina moved a sailboat so that its mast blocked the beam.

fy20• 3 years ago

Point to point WiFi links are pretty accessible nowadays. I bought a €60 no-name set from Amazon to link two buildings 250m away that don't even have line-of-sight (other buildings and trees are in the way) and it works good enough for my needs (CCTV).

robinduckett• 3 years ago

Two university buildings separated by a privately owned house or two springs immediately to mind. Not sure if they still do but my local university used to have IR links between buildings.

readams2• 3 years ago

They're not talking about point to point links here, but end-user devices.

lxgr• 3 years ago

Signals not propagating through doors or walls can be an advantage for many applications, both for capacity and privacy reasons.

flashback21992• 3 years ago

I'm imagining offices that put Li-Fi devices on the ceiling and laptops made with Li-Fi in the edge of the screen lid so that you can just open your laptop and it will start using the Li-Fi for faster speeds.

haswell5• 3 years ago

This would be fantastic in my current apartment. My immediate surroundings are completely saturated by WiFi access points across both 2.4 and 5.x GHz making all wireless connections unreliable at best, and the layout of my apartment is such that the only way to have a wired connection in the living room where my desktop sits is to run an ethernet cable across the floor.

So I run an ethernet cable across the floor, and it's not great.

I also think VR/AR use cases may benefit from this eventually.

wyldfire• 3 years ago

I wonder, does this pollute the IR on your apartment so that your TV remote would no longer work? Although with streaming set-top boxes, those are using RF most of the time these days.

jorvi2• 3 years ago

It is virtually impossible you’re getting enough distortion from other 5GHz networks to matter, unless your walls are made out of tissue paper. I would recommend analyzing your own devices for faults.

rsynnott• 3 years ago

You could consider wifi6e; the 6GHz band probably isn't too noisy, for now.

pizzapill2• 3 years ago

Have a look at Ethernet over power cables. I had them in my last apartment where WIFI was bad and cable was unacceptable.

themoonisachees• 3 years ago

Someone mentionned VR headsets (if they can actually make the speeds advertised work)

Office spaces are going to eat that shit up. They are already running Ethernet everywhere, and they already have many wifi access points. This is probably lower-crosstalk, faster and more localized than wifi.

letitbeirie2• 3 years ago

This is pretty cool - being able to identify/segregate traffic by its physical room seems like it would be a useful tool for anyone in security, university professors giving exams, etc.

I can't really see it totally replacing WiFi though. It's hard enough to get a TV remote to work outside in the shade - getting higher-bandwidth IR-based WiFi to work outside in the sun seems... ambitious?

dragontamer4• 3 years ago

IrDA above 115kbps is basically dead, and I don't think anything beyond 4MBps is feasible (ie: months spent on an FPGA to handle all the issues)

I'm a big fan of using simplified IrDA in 2023 in hobby electronics. But I don't see a smooth path to 9.6Gbps.

1Mbps to 10Mbps is already exhausting for me to think about.

--------------

Is there any industry chip with planned development to handle this standard? Finding a damn LED and Photodiode that is guaranteed faster than 20ns to feasibly handle 4Mbit is horrifying. Let alone the signal processing circuits (I think you need 20Mhz OpAmps which isn't a 6001 or lm358 if you know what I mean) or FPGAs needed to handle this relatively low speeds.

Or are we doing some kind of multi-frequency multi LED thing? Are there any LEDs or Photodiode that can emit and detect such a narrow band? What are the sources of IR noise to mess with those frequencies?

-------

Oh, but 115kHz is plenty useful for a hobbyist and I highly recommend people playing with that.

zokier3• 3 years ago

I suspect that IEEE 802.15.7 might be more up your alley, maybe also .13 (esp the pm phy variant).

flashback21992• 3 years ago

Edit: I re-read your comment and I think I see what you might have meant now. I think you were asking if there would be a solution for low speed devices to use this standard.

dragontamer4• 3 years ago

I'm asking if the LEDs that support 9.6Gbps (what is that? A 1ns or faster LED?) even exist in any substantial way.

Isn't that like, laser speeds? Not overhead LED speeds. And over the air? Not inside a fiber line?

This is basically a 10Gbps standard for fiber lines except over the air. All those issues need to be solved without the guarantees or isolation that fiber optic lines provide.

That's just transmit. Receive has to deal with the same issues, amplifiers and other electronics that move at the requested speeds.

_trampeltier• 3 years ago

There are much faster IR Links for industrial application. They are for example used in automatic crane application. I don't know the data rate just now, but we have a crane with a such Profibus link since over 20 years in our factory.

https://www.leuze.com/en-us/products/optical-data-transmissi...

dragontamer4• 3 years ago

Well... I think I can visualize a 100 Mbit IR link in theory.

If we use Wi-fi like processing (256-QAM and the like), we'll only need 12.5 Mhz-class LEDs and photodiodes to communicate at 100 Mbit. (assuming 8-bits per 80 nanosecond or whatever). I'm pretty sure that cheap LEDs with 50 nanoseconds aka 20MHz speeds, as well as analog components with high-gains at 20MHz (I dunno, a 1GHz bandwidth-gain, giving x50 gain at 20MHz) are all possible.

I think the article's claims of 9.6 Gbit are so absurd however, that I'm having a great amount of difficulty seeing how it gets there.

I admit that I wasn't aware of the automation IR links that you posted. I'm impressed that IR has gotten that good today, but its not so good that its become unbelievable. The 9.6 Gbit headline on this article however is just absurd at face value. I really need to see a physical demo before I'll believe something like that.

-------

I know that IrDA is a 90s standard made for far slower LEDs and components. But its a good baseline for thinking about the issues that occur in the space of light-based communications.

eternityforest• 3 years ago

I'm sure once the demand is there, the silicon will be mass produced.

I like 38khz modulated IR more than 115kbps, you can get really good range at low cost and power, without high parts count, and if you have large amounts of data rather than just stuff that takes advantage of directionality, you can do it over Bluetooth for cheap.

Never got a chance to play with 115kbps though, seems like it could be pretty cool, but I'm not sure what I'd do with it.

RobotToaster• 3 years ago

As someone pretty sensitive to CFL/LED flicker, the idea of this being everywhere is worrying.

newaccount74• 3 years ago

LED flicker is 50 or 60 Hertz, this technology would be Giga-Hertz. And it's infrared. Your eyes won't see a thing.

mrob• 3 years ago

50 or 60Hz is low enough that even people who aren't sensitive can notice it. Cheap LED lighting more often flickers at 100 or 120Hz, using a bridge rectifier with a series capacitor to limit the current.

hulitu3• 3 years ago

https://en.m.wikipedia.org/wiki/Stroboscopic_effect

fatfox• 3 years ago

it’s invisible

ksec• 3 years ago

Oh FFS. So they approve 9.6Gbps Li-Fi aka IEEE 802.11bb, and the devices shown in the photo used inside a prototype is only capable of 1Gbps, but every article including one submitted last week said "100x Faster " [1] than WiFi at 224GB/s or 1792Gbps.

And just like 802.11be and 802.11ax, I had to read up on the actual Spec again to understand WTF is going on instead of relying tech reporting. And this isn't just a problem with reporting, the actual PR of Li-Fi cant even do half its job.

[1] https://news.ycombinator.com/item?id=36707789

sp1rit• 3 years ago

I remember my old Nokia brick having the ability to send files and apps through an infrared thingy on its side. You could simply hold two phones side by side from each other and transfer data, but I remember handly anyone using it (I guess it was kind of akwad to use - but probably still simpler than getting the devices paird with bluethooth).

It wasn't nearly as fast the advertised speeds mentioned in the article.

spdustin• 3 years ago

I’d be curious to know how it impacts things like FaceID. Or if bathing a room with rapidly modulated IR would affect thermography (as used in, say, energy efficiency audits).

wiradikusuma• 3 years ago

"..instead of having wifi that someone can tap into.." so, use Ethernet cable?

throitallaway2• 3 years ago

A lot of this does feel like reinventing the wheel. However, it's slightly more convenient than being tethered with a wire.

gochi2• 3 years ago

Sure let me attach an ethernet cable to my VR headset, and then magically pull another from the wall to attach to my laptop. Can't forget the mystical third wire which I need for my phone!

Hopefully it doesn't take much to grasp the benefits of ethernet without a cable directly to your device at all times, right?

DaSHacka• 3 years ago

But those devices, particularly the VR headset and mobile phone, are not good candidates for an IR-based networking solution as their use-case inherently has a lot of movement. What good is a VR headset that loses connection with the host computer when you accidentally block the receiver while moving around?

deepnotderp• 3 years ago

One day the use of self-accelerating beams (which can bend/turn by themselves) may remove the line of sight limitation.

https://www.mdpi.com/2076-3417/7/4/341

jimmytidey• 3 years ago

I remember reading that Royal Navy engineers maintaining nuclear submarines use Li-Fi to connect their workstations to the network.

They said it was a security requirement - does seem extraordinarily elaborate. It feels like using wires could have been a simpler answer...

jchanimal• 3 years ago

Wires likely permit too much side-channel to be safe.

throitallaway2• 3 years ago

Interesting and agreed. Is it more secure (or less detectable) than Ethernet? I suppose Ethernet has some form of magnetic radiation. I wonder what all this infrared light bouncing around will do to us monkeys.

woah• 3 years ago

Make us infinitesimally warmer?

mcpackieh• 3 years ago

I don't mean to alarm you, but you emit infrared light.

hparadiz• 3 years ago

If I have a room full of these can they communicate with each other to provide a connection? This reminds me of base stations for my VR setup where my feet trackers will sometimes lose connection due to a physical obstacle like a desk. The solution is to add more base stations typically or make your space more sparse.

gochi2• 3 years ago

They ran a case study at a school, you can scroll down and see their overview setup which did what you're talking about, a room full of them providing a connection: https://www.purelifi.com/case-studies/lifi-in-the-classroom/

njanirudh• 3 years ago

Is https://youtu.be/Rw-QZuGVP3o somehow related to the technology in the article.

blincoln2• 3 years ago

Seems like the fact that it's a "tightbeam" vs. WiFi's broad coverage could result in some fun niche hacks like tiny autotracking IR transceiver turrets that follow devices around a room or (working together to hand off the connection) a building, or larger versions that keep two transceivers pointing at each other over a longer point-to-point link. Who will be the first hobbyists to establish a LiFi link bounced off of one of the old retroflectors on the moon?

If the IR is bright enough, it might even make sense to put a NIR camera in every access point so it can double as an unobtrusive security camera, and/or track devices (for beam-forming, forensic information for physical security staff, etc.).

jdblair2• 3 years ago

No mechanical turrets needed! Just use a flat, stationary, phased array infrared emitter.

nuccy• 3 years ago

Do you have a reference on any of those? I can only find papers about lab setups. Since wavelength of IR is much shorter phase control of different emitter is much harder to accomplish than in the radio. Moreover thermal deformation of the device would affect the phases much more significantly.

Animats• 3 years ago

HP used to sell an infrared LAN. Little red ceiling domes in offices. It wasn't popular, but it did work.

Havoc2• 3 years ago

Given how much people struggle with vanilla routers I’d say anything requiring line of sight could be challenging.

I could see it being supplemental though. Wifi and this sorta like anycast

firefoxd• 3 years ago

Not to bash on this tech, but at the close proximity and fixture it requires, why not a wired connection?

atum47• 3 years ago

There were some research in the past where they can turn WIFI signal into video (seeing through walls, detecting humans...).

I wonder how easier this will make things for them...

6510• 3 years ago

Perhaps it would even be useful to do networking use audio (like dial up sounds or cassette tapes) One could carefully pick sounds to be the least annoying and have [say] a printer announce to the world that it is still online. One could probably transfer well over 50kb/s but it hardly seems necessary if one rarely prints anything. I already scream at rarely used printers it would be great if it would listen for print jobs. It could take dictation too. Perhaps it is funny to have the entire print order in human speech.

Iphones have some weird feature to listen for coughing or water(?) We already have door bells and door knockers sending out audio signals. (Video doorbells are great but knockers are more reliable.) Lots of devices in the home make identifiable sounds. Combined with the security footage you could generate a film of you using your wireless drill, every time you look in the fridge, open the front door, when you shoot someone, etc "vacuum cleaning the movie".

The vacuum cleaner can be asked for a status report, its serial, build date etc, how full its bag is utilizing a human friendly machine readable audio signal.

It seems much more secure than evil men silently communicating with your devices over the internet.

I think we are onto something here.

elcritch• 3 years ago

Where I believe LiFi options would be great is to eliminate HDMI / Display Port cables to monitors or TVs. Just one less wire to run and figure out how to hide. Well at least until wireless energy systems get better. ;)

Or to have a desktop computer across the room from a display and be able to switch displays.

kbumsik• 3 years ago

I wonder how it offers bi-directional communication?

Do mobile devices have to have small light bulb (like infrared) to support bi-drectional communication? It would be too big for modern smartphone though.

Infrared may no longer be a punchline, as IEEE approves 9.6Gbps wireless light

Related Stories