Radical new idea
Oh, hang on...didn't my old mobile phone about 15 years ago have an option to communicate via infra-red? OK, not at 43Gbps maybe...
And aren't there line-of-sight issues?
In five years or so, Wi-Fi access points could carry data at rates 100 times faster than today using infrared light rather than other portions of the electromagnetic spectrum. Researchers at Eindhoven University of Technology (TU/e) in the Netherlands have developed a way to use fiber optic cables, mounted with networking …
IrDA and the like has indeed been around forever.
And site-to-site infrared / microwave systems too.
They have serious problems that stop them being used for anything other than consumer toys (IrDA way surpassed by Bluetooth, line-of-sight kills anything but use over your own land or free space).
Been trying to convince my employers to put in site-to-site-links but the only places that can "see" are in positions that would show the ugly, and they disapprove.
Instead, we spend £10k a year on a Virgin leased line and another £10k on a BT leased line at the other end (because no one supplier covers both areas... sigh). When I can see each building from the other. Just a shame we don't own all the land in-between because I'd just get them to chop down a couple of trees and problem solved.
We had IR comms before.
It's amazing how the sun rising in the morning can really balls things up...along with pigeons, leaves and people changing the street lights on a cherry picker.
If the link ever went down, it was standard practice to get on the roof with a pair of binoculars and see what was in the way.
I might be wrong about these things, but I thought that if you modulate a given signal with a given bandwidth onto a carrier frequency you will occupy the spectrum around the carrier frequency equal to the bandwidth of the original signal. If I am correct about that, will it not be so that there is a lot more room around the higher frequency bands? So while it is true that higher frequency does not dictate higher data rates, it does indeed allow for it.
Take for example a signal of 1 MHz bandwidth. Transmitted on 2MHz it will occupy the area from 1.5MHz to 2.5MHz, a rather large part of the spectrum, effectively blocking out anything else from being transmitted at the same time. Shift it up to 1GHz on the other hand and you will occupy the spectrum from 0.9995GHz to 1.0005GHz, allowing for other signals to be transmitted nearby, or expanding the bandwidth of the signal in the first place.
Disclaimer, I am unsure if I remember these things correctly so please correct me if I am wrong.
that moves around and emits a whole shed load of IR through to UV and beyond?
Namely the Sun.
In a darkend room or deep underground etc yes this might work but outside in the real world with the Sun and a whole host of other IR light sources this would at first glance have very limited applications.
Perhaps after the Yellowstone Caldera blows its top and the dregs of humanity that are left and have to live without the sun and clear skies dor possible decades then this might work.
Skeptic is my middle name.
"...that moves around and emits a whole shed load of IR through to UV and beyond?"
Seriously? Have you never wondered how IR remote controls manage to work in the daytime when there's that 'thing in the sky', even though you may not have closed your curtains or blinds?
Could be handy to get a network up and running to get a revenue stream in place before a more permanent solution is put in place or where a fast temporary fix is needed say, across a major bridge/road, where a cable has been damaged, and there would be inherent dangers, using other methods.
Could help in the rollout of rural broadband, but I doubt it will be cheap, as in as cheap as stringing a temporary fibre cable between two points.
If Germany goes ahead with its Euro 100Bn investment in Fibre / 5G + other Countries, fibre itself is going to get a whole lot cheaper over the next 5 years, aswell more refined, cheaper tried and tested methods of installation.
If you can attach a seemingly-harmless IR transmitter to someone's network, you can transmit content to a remote location within line of sight . . . for example from one colocated data center cage to another or from one high-rise, glass-enclosed building to another. Given that it's point-to-point line-of-sight, it should be largely undetectable by conventional means. You could potentially also use it for regular networking under similar conditions, where you have physically proximate but separated sites. There could also be use cases where RF interference is not desirable but wireless connectivity is. Not everyone will find the technology useful, but there do seem to be specialist applications possible for it.
Who the heck needs 42 Gbps wifi? Heck, who really needs 420 Mbps wifi? Same as the objection for who really needs gigabit broadband to their home - no one except a narrow portion of people who work with truly massive files, like CAD engineers working from home. We have maxed out the input our senses can handle with high quality streaming video, and even Blu Ray quality 4K is only 100 Mbit (as if we'll ever be getting that level of quality from Netflix or Amazon Prime) The problem people have with wifi today isn't the 3 Gbps theoretical speed, it is that it is like 1-2% of that speed when you have to go through a few walls. Infrared requires line of sight, so it won't help you there.
I could see this being useful for lets say an army brigade during war sets up a new HQ, and has a bunch of computers. Rather than running fiber everywhere in a room that wasn't designed to be a datacenter, just put like IR "antennas" on top of each rack.
I wonder if an IR base station would drown out your ability to use the remote control for your TV?
Who the heck needs 42 Gbps wifi? Heck, who really needs 420 Mbps wifi?
Well for one use case, when I'm doing backups at home I try and plug the laptops into a wired network - it makes a VAST difference to backup speeds. It's not just the raw speed, it's the latency as well - there's a lot of round trip exchanges in determining which files needs copying.
If your backups took 5 minutes instead of 50 minutes, what difference would that make to you? You aren't waiting on them, they're taking place in the background. Sure, better round trip latency would be nice but TCP ACK packets are tiny, and a lot of that latency is in the network stacks of the endpoints, so you won't improve your latency by much at all via speeding up the link.
As far as stuff YOU have to wait on, or things such speeds would enable you to do that you can't do now, what is there? I ask this question all the time, and aside from the unimaginative folks quoting "640K should be enough for anyone" at me, I've never heard anyone come up with a reason why we need more than a gigabit into the home, or between any two devices in the home.
The reason we've demanded more and more speed is because it has enabled us to spend less time waiting on our computer for e.g. web pages to load, or to get denser sensory data (i.e. going from text to GIF to MP3 to 2005 era YouTube to 4K streaming today) but we've maxed out the richness of input possible to our senses.
Going from a gigabit to 10 gigabit, for instance, the loading time for a 5 MB web page would improve from 50 ms to 5 ms - while you could notice that, it wouldn't improve your experience any. Of course, we all know such pages don't load in 50 ms on a gigabit link, or even 500 ms. Because of all the back forth traffic, slow ad servers and what not it takes seconds. A faster link won't make any difference.
So aside from hand wavey "maybe someday someone figures out how to do a Star Trek like holodeck, which takes 10 Gbps" are there any real world use cases that demand faster than a gigabit? Things that make a measurable difference in your user experience, not "if I set down my laptop right before bed, it can finish its backup before I even fall asleep, rather than not finishing until I've entered my first REM sleep".
People are always resistant to the idea that we've hit a wall for homes (well not those who are still at 50 Mbps or whatever, but places where you can get a gigabit) where faster speeds aren't needed, and there's no reason for 10 Gbps ethernet to come to homes, either. They've been so used to networks getting faster and faster, and more data being used, that they assume it will continue forever. It won't. There has to be a use case driving it, it isn't "build it and they will come".
The same thing happened with PC CPUs almost a decade ago. Everyone was used to CPUs getting faster and faster, and assumed they always needed a faster CPU, until they pretty much stopped getting faster. How much faster are today's fastest PC CPUs compared to ones almost a decade ago with the first "Core i3/5/7" CPUs? Not even close to twice as fast in single thread CPU. They are still quad core at the high end of typical desktops today, just like they were back then. But it doesn't matter - if you could get a CPU 5x as fast as what you have now, but had to give up your SSD and go back to a hard drive, unless you're an idiot, you'll keep your SSD.
Rephrase it another way Doug, what problem does this cause you that you can write paragraphs and praragraphs rebutting the usefulness of it. Why *shouldn't* we have faster wifi? Can you think of *no* situation where faster wifi would be an economic benefit?
Here's a hint: its not always about home networking, individual computers or loading a web page. There are people out there who do different things with networks than you! Please don't faint in shock.
"if you could get a CPU 5x as fast as what you have now, but had to give up your SSD and go back to a hard drive, unless you're an idiot, you'll keep your SSD."
So why did we develop SSD's surely a 7200rpm disk is way more than a consumer needs. You never need enterprise speeds of 15K
If your backups took 5 minutes instead of 50 minutes, what difference would that make to you? You aren't waiting on them, they're taking place in the background.
Well it's actually the difference between fraction of an hour and "hours" - so yes it is a significant difference in time.
And yes I am waiting on them. At present they tie up my laptop - I can use it, but while it's running backups it's "a bit sluggish". When I manage to clear the significant list of things that require a dip into my "far from infinite" income then I'll be able to improve things (either a laptop that can have much more memory than this one supports, or a separate machine to run backups on, or both) - but at the moment that's not a "next week" timescale.
The other issue is how long the machine being backup up is on for. Sometimes it can be difficult persuading people to not shut the machine down before it's backup is finished.
There's no consumer use case for this
Sure there is. I would not have had to run a GigE in conduits to my workshop/shed. That is my "off-site" backup so WiFi just doesn ot cut it - you get at most 30-40Mbit at that distance which is just not enough to shovel the 40G or so of daily backup on a daily basis.
I get that This Just Sounds Outrageous.
I just spent a week getting all the details straight so we could get our new rack cabled and fibred up so we could spin up the hosts in that rack. Three of those days were spend just plain running cables/fibres between switches.
Is it not possible that say, having this wired up in the DC and connecting endpoints or even TOR switches to EOR switches might be worthwhile?
Hats. Better to use one made by a milliner rather than the one yer sitting on.
"In five years or so, Wi-Fi access points could carry data at rates 100 times faster than today using infrared light..."
This is a great idea, so long as the link is inside a building or if there isn't any snow or fog. Many decades ago, a similar link (not nearly as fast) connected the CU Boulder Engineering Building with the Computing Center approximately 2 miles away. Worked great until we had snow. Or rain. Or fog. The link completely dropped out then. IR doesn't like that. Sun didn't seem to matter, though.
Also, multiple sources pointed at a single IR transceiver will screw up. A (now long-dead) HP project used IR for a wireless hub. One serious problem was sources who were closer swamped traffic from those further away. HP never totally figured out the problems and so never released the product.
But boffins are so much cleverer these days! Nothing can go worng <tick> go worng <tick> go worng <tick> go worng <tick> go worng...
Cat videos at lightspeed - check
Pr0n at lightspeed - check
UHD pr0n at lightspeed - check
UHD 3d pr0n at lightspeed - check
Multiple simultaneous streams of holographically enhanced super UHD 4d cat pr0n at lightspeed - checking - er - er - ....
... Welcome to our new 802.11x. enhanced xray wifi overloards.
Hey, only 2 letters left. Time to develop a new standard. Anyone know the speed of a pigeon in a vacuum?
Biting the hand that feeds IT © 1998–2019