Wasn't this a completely different story a few hours ago?
Japanese geniuses have maintained a 3Gbit/s radio link at 542GHz, opening up more of the electromagnetic spectrum to the voracious appetite of wireless data. Not that the research will lead to super-Wi-Fi any time soon - but if the proof-of-concept tech built by the Tokyo Institute of Technology can be commoditised then it …
Wasn't this a completely different story a few hours ago?
The old one's still there, just not on the index page.
Well, that's what some people think which is why they want the T-hertz scanners banned.
And don't spare the horses.
... then why not just use infrared or visible light to start with and have an even higher throughput?
Or am I missing something?
Cloud, rain, fog, smog, dust clouds, flocks of birds etc causing intermittent or continuing communications problems. Also people and wildlife damaging their eyesight by looking into the beam. Limited range. Problems with opaque material building up on the emitter, requiring it to be cleaned periodically.
Do a search for Free Space Optical networking, and for a nice open-sourcey example, RONJA. Basically, its only useful when you can't use wired or radio networking, and that's a fairly limited market these days.
"Cloud, rain, fog, smog, dust clouds, flocks of birds etc causing intermittent or continuing communications problems"
Not major problems inside a house. Well, I suppose if you're a chainsmoking Dr Doolittle...
"Also people and wildlife damaging their eyesight by looking into the beam"
Doesn't have to be a laser. Non lased light can do the job just as well using fast switched LEDs.
"Basically, its only useful when you can't use wired or radio networking, and that's a fairly limited market these days."
For outside sure , its probably a non starter. But for across a room there should be an issue.
You mean like your TV remote?
Fast switching of LEDs is how TOSLINK optical and TV remotes work. The bandwidth is very low as it can only use brightness for data transfer, because the LEDs emit relatively wide band, unpolarised random phase radiation.
That also makes them extremely reliable in terrible conditions.
In proper fibre optics the bandwidth is much higher because the laser diodes are extremely narrow band and in phase - if not polarised as well. So much more possibility for data transfer.
Which of course means they need very tightly controlled conditions - the inside of a glass fibre.
"For outside sure , its probably a non starter. But for across a room there should be an issue."
If you're only doing cross-room stuff, then modern 60ghz+ radio stuff would be absolutely fine, no? I thought the article was about longer ranged outdoor transmissions, which are a rather more tricky prospect. Indoors stuff though... I'd want to use radio, because my house is full of things which are opaque to visible, near and far infrared light. Some sort of IR strobe that bounces a signal off the walls sounds like it would just cause all sorts of crazy multipath degraded signals.
"Doesn't have to be a laser. Non lased light can do the job just as well using fast switched LEDs."
Oh sure; RONJA uses LEDs. That doesn't mean you'd want to stick your head under the rain hood, or peer at the emitter with a pair of binoculars to see if it is turned on.
Have a look at VLC - Visible Light Communication http://visiblelightcomm.com , there's a research group at Edinburgh doing clever modulation of while LED lights to perform the same function as WiFi (the term LiFi seems to have been adopted at some point).
One of the professors responsible did a TED talk: http://www.ted.com/talks/harald_haas_wireless_data_from_every_light_bulb.html
It's not 1.5GB/s but it can stream video already, one of the current projects is a tiny ultra low power modem to accompany the light in it's fitting (with an ethernet connection probably), rather than the bulky prototype they have just now.
We already knew we could use this spectrum, but would deal with these issues. Nothing new to report here except some people want some funding to continue their research into tech which can only replace NFC. This will never replace WiFi because to get the same penetration & range you would have to turn up the power to the point where you'd cook the birds in the sky or something. With better antennae it is easy to boost the range of existing tech. I'll keep my homemade point-to-point 802.11g(108Mbps) link thank you.
"which stretches up to 400THz and then we're into visible light and talking lasers"
Where can one purchase one of these talking lasers? Is it like siri?
Since the beginnings of radio, RF engineers and scientists have continually tested the upper frequency limits of the latest technology. And this account says the state of the art is 542GHz--or it's at least the point where a signal can be modulated to carry useful information. Using WWII as a convenient reference point (as practical commercial/military comms where f=10GHz (lambda 3cm) had been developed), then it's interesting to note that progress in 70 years has increased the upper RF frequency limit by about 54 fold!
As boltar says, these frequencies will be pretty useless for WiFi as we know it. They'll be stopped by just about anything including refraction by the atmosphere (and at close quarters, where levels above a few milliwatts would be expected, the radiation would be dangerous, especially to the eyes.
In effect, what these researchers have done is to modulate a (very) 'longwave' heat signal.
[You heard it here first.] Where these extreme frequencies would be useful would be to provide very wide bandwidth communications between say the moon and Mars--there's no atmosphere in between except at Mars (and that's thin), or between interplanetary satellites. There'll be little or nothing to stop or refract the waves and the power output can be whatever it needs to be w/o the possibility of injury to anyone.
Just done a quick calc. 542GHz = ~0.5535mm wavelength, whereas heat (IR) from the sun spreads very roughly over a wavelength of 800 - 2000nm. At 2000nm, the IR from the sun is about 10% or less than at its maximum point.
The sun's radiation closely approximates aperiodic black body radiation peaking at about 550nm, thus if we use 1000nm as the reference point (that's definitely in the IR region) then the frequency of the radiation calculates out at about 300,000GHz. That's ~554 [3x10^5/542] times the freq. of our 542GHz comms channel.
The path loss across these interplanetary distances will be very high even with extremely directional antennae (which are easily constructed at these frequencies), so it's important that there be a sufficiently high protection ratio from the sun's thermal radiation. As the enormous heat from the sun 'centres' some hundreds of times higher in frequency than our communications channel, we'd hope that good IR cutoff filters, directional antennae etc. would ensure interfering noise from the extremities of the sun's spectrum---the bandwidth edge--would be low enough so as not to swamp the 542GHz comms circuit.
Even w/o taking account of any interference from the sun, the circuit's equivalent thermal noise has to be low enough to enable communication over these distances. Noise is always a big issue in communications circuits, especially so at bandwidths as wide as 2GHz. Thanks to the ubiquitous thermal noise equation, vrms = (4ktbr)^0.5 and variants, it's easily calculated.
Ok, so this my instant back-of-a-postage-stamp calculation. For those of you who've detailed/current data about space communications, it's now up to you to calculate the TX output power for a Moon/Mars circuit with a fade margin of say 40dB (i.e. that of a good commercial-grade microwave comms circuit).
Please folks this is all farting around at the edges. The only answer is FTTH and then use you wifi tech of choice. A single optical fibre can currently transmit the 300,000 times the bandwidth of ALL the radio spectrum. Bert