Slashdot Mirror

← Back to Stories (view on slashdot.org)

Scientists In Japan Build 100Gbps Wireless Network Using Terahertz Transmitter

Posted by timothy on from the stop-hurting-tera! dept.

Mark.JUK writes: A group of Japanese scientists working on a project managed by Hiroshima University claim to have successfully built a Terahertz (THz) transmitter, which is implemented as a silicon CMOS integrated circuit and can transmit a signal running at 10Gbps per data channel over multiple channels in the 275-305GHz band for a top speed of 100Gbps (Gigabits per second). But crucially nobody has mentioned the distance at which this speed could be achieved, particularly since the THz band isn't likely to have much of a reach. It also sits very close to the region used by lasers.

49 comments

  1. ^^ Orly? by Anonymous Coward · · Score: 0

    Maaaaaximmooooom Poowaaaaaaru!!!

  2. Huh. by pushing-robot · · Score: 1, Insightful

    TIL lasers have their own band in the EM spectrum.

    --
    How can I believe you when you tell me what I don't want to hear?
    1. Re:Huh. by Anonymous Coward · · Score: 0

      What kind of lasers? Frickin' lasers?

    2. Re:Huh. by radarskiy · · Score: 1

      That band is called "light", hence the 'l' in laser.

  3. Lasers at 300 GHz? by Anonymous Coward · · Score: 0

    Which laser operates at 300 GHz? This is still firmly in the world of RF radiation. Most lasers operate in the IR or visible light range which is more like 10000GHz, not 300. The article also mentions some clickbait bullshit about health concerns because of this (nonexistent) laser operation region, there is no evidence THz radiation is bad for you - it is still well below the threshold for ionising effects. Not to mention the fact that lasers are only dangerous because of their high intensity, they are not inherently dangerous.

    1. Re:Lasers at 300 GHz? by TechyImmigrant · · Score: 2

      >Which laser operates at 300 GHz?

      Masers
      .

      --
      I should use this sig to advertise my book ISBN-13 : 978-1501515132.
    2. Re:Lasers at 300 GHz? by Anonymous Coward · · Score: 0

      It's not really a Laser then if it's a Maser.

    3. Re:Lasers at 300 GHz? by TechyImmigrant · · Score: 1

      It's not really a Laser then if it's a Maser.

      Same principle of operation, but it's reasonable to expect light to come out of a laser, yes.

      --
      I should use this sig to advertise my book ISBN-13 : 978-1501515132.
    4. Re:Lasers at 300 GHz? by Khyber · · Score: 0

      Microwave lasers. EM can still be directed and made into coherent beams no matter the wavelength. Wave-particle duality FTW, son. Go back to basic physics class in high school. Oh, wait, you're likely one of the NCLB generation.

      --
      Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
    5. Re:Lasers at 300 GHz? by Anonymous Coward · · Score: 0

      Why do you feel the need to throw insults? It doesn't make you look intelligent, it makes you look like a Donald Trump.

  4. One Word by Anonymous Coward · · Score: 0

    Soylent. Green. Is. Made. Of. North. Korean. Labourers.

  5. How close to the lasers' region? by davidwr · · Score: 1

    It also sits very close to the region used by lasers.

    Within firing range?

    --
    Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
  6. When did 300GHz become Terahertz... by Anonymous Coward · · Score: 0

    So, I always thought SI units specified that 1 THz = 1000 GHz. According to the article, 1THz is 300GHz+. This is obviously some new math I've never heard of.

    1. Re:When did 300GHz become Terahertz... by Phreakiture · · Score: 1

      Yes, I agree with you. The only possible explanation that I can come up with is that they must have dubbed the next order of ten over EHF as "The Terahertz Band". If that's the case, it needs a better name. If we are to assume that they are keeping with the general pattern used for radio signals, then this band would go from 300 GHz to 3 THz.

      This would be an extension of the standard used to class radio waves into bands of an order of ten, which gives us VLF (9-30 kHz), LF (30-300 kHz), MF (300-3000 kHz), HF (3-30 MHz), VHF (30-300 MHz), UHF (300-3000 MHz), SHF (3-30 GHz) and EHF (30-300 GHz). VLF is truncated because EM waves below 9 kHz aren't considered radio. The lower ones are sometimes also called by wavelenth rather than frequency, so VLF-HF may also sometimes be called VLW, LW, MW, SW, though these are usually spoken full-length, e.g. "short wave".

      Back on topic, I'll repeat: it needs a better name. EHF stands for "Extremely High Frequency" . . . what's more extreme than extreme? Hyper, maybe? HHF?

      --
      www.wavefront-av.com
  7. wtf indeed by Anonymous Coward · · Score: 0

    and I learned that 350 Ghz is a Terahertz

    1. Re:wtf indeed by Anonymous Coward · · Score: 0

      https://en.wikipedia.org/wiki/Nyquist%E2%80%93Shannon_sampling_theorem

    2. Re:wtf indeed by Anonymous Coward · · Score: 0

      Yes, thanks for the link. Now what does it have to do with the story?

    3. Re:wtf indeed by Khyber · · Score: 1

      That sampling theorem got blown out by any real audio technician that uses pitch-shifting.

      That theorem is NOWHERE CLOSE to being adequate for any situation where pitch-shifting is a requirement. You need at least 32x the sampling rate to achieve near-flawless pitch-shifting up or down to a MEASLY TWO WHOLE NOTES, NO MATTER WHAT THEIR FREQUENCY RANGE.

      I've proven it with .FLAC files, .RAW files, .WAV files, and .PCM files, across every pitch-shifting plugin known for the most popular media players.

      Shannon-Nyquist are old-age. Time for a new theorem.

      --
      Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
    4. Re:wtf indeed by Anonymous Coward · · Score: 0

      You are insane. That's like saying Newton is "old-age" [sic] because Boeing makes airplanes. You don't understand what you are talking about.

      Judging by your posting history, *no one* does.

  8. Nyquist can stop spinning. by SETY · · Score: 0

    TFA says "and now another team claims it can hit 100Gbps by pushing into the TeraHertz (300GHz+).". Nyquist can stop spinning.

    1. Re:Nyquist can stop spinning. by Plammox · · Score: 1

      Did you ever hear about quadrature modulation?

    2. Re:Nyquist can stop spinning. by Khyber · · Score: 1

      That doesn't actually preclude physics. We can get multiple bits of information from a single photon. QAM, frequency, phase-shift, etc. The theorem still holds across almost every perceivable 'sound effect' until you get to the matter of pitch shifting, in which case, even with the minimum 2x sampling rate, the reality is that you need about 32x that amount to get an accurate signal representation that won't distort across a wide range of frequency shifts.

      There's a reason Nyquist-Shannon has a "THEOREM" instead of a "LAW."

      --
      Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
    3. Re:Nyquist can stop spinning. by Anonymous Coward · · Score: 0

      That doesn't actually preclude physics. We can get multiple bits of information from a single photon. QAM, frequency, phase-shift, etc. The theorem still holds across almost every perceivable 'sound effect' until you get to the matter of pitch shifting, in which case, even with the minimum 2x sampling rate, the reality is that you need about 32x that amount to get an accurate signal representation that won't distort across a wide range of frequency shifts.

      There's a reason Nyquist-Shannon has a "THEOREM" instead of a "LAW."

      Baloney. Nyquist-Shannon comes with a PROOF. That actually makes it stronger than an empirical law.

      Your pitch-shifting may need to internally process the data at a sampling rate 32x the highest component, but you don't need to sample at that rate in the first place.

      You can sample at the Nyquist rate and interpolate/reconstruct to any higher rate internal representation that makes manipulation easier. And you can decimate again after applying your effects, again without losing any information.

      Moreover, this is not just theoretical. Have a look at e.g. the MATLAB `upfirdn` filter implementation.

    4. Re:Nyquist can stop spinning. by Anonymous Coward · · Score: 0

      You are very courageous. Khyber is a howling lunatic with no grasp of what he's talking about, and as well-intentioned as you may be, it's futile.

      And " get multiple bits of information from a single photon"??? This idiot is worse than a child having a tantrum.

  9. real world testing by funkymonkjay · · Score: 1

    throw in a running microwave oven and at a distance of 1/2 mile.

    1. Re:real world testing by Phreakiture · · Score: 1

      Wouldn't even register. 2.4 GHz is so far out of the tuned range it's not funny.

      --
      www.wavefront-av.com
  10. Yes But? by JimSadler · · Score: 1

    It sounds wonderful but will it cook my turkey if I leave it on the window sill?

    1. Re:Yes But? by Phreakiture · · Score: 1

      No.

      That is all.

      --
      www.wavefront-av.com
  11. Why is this a network? by avandesande · · Score: 1

    Maybe this is more useful for direct data links? Might be a good way to send data across the ocean.

    --
    love is just extroverted narcissism
    1. Re:Why is this a network? by Anonymous Coward · · Score: 0

      It will be useful for point-to-point and local-broadcast links; Lower frequency microwave systems have been used for such p2p links for a long time.

      Unfortunately it, like anything else higher than roughly 10-20MHz, is useless for over-the-horizon communication because everything above roughly 20Mhz goes through the ionosphere instead of reflecting off of it.

  12. This looks familiar by Plammox · · Score: 1

    I seem to remember a publication in Nature from 2013. Seems the Germans already did something comparable a few years ago. Over 20m distance that is.

  13. 100 gbps isn't that fast by SmaryJerry · · Score: 1

    4G has offered these speeds for years between cell towers. For individual towers, it gets split among all users and phones, which also don't use enough power to upload at the same speed and range as the tower. But a THz transmitter sounds interesting. Anyone know if the propagation would be better/worse? I would imagine worse?

    1. Re:100 gbps isn't that fast by Mad+Bad+Rabbit · · Score: 1

      Terahertz signals would only reach 10's to maybe 100 meters, AND they're blocked by metal or water.
      (i.e. a human standing next to the router might cause significant signal loss). Probably good enough
      for indoor use, though --- at 100 gbps you could afford to retransmit lost packets.

      --
      >;k
    2. Re:100 gbps isn't that fast by jratcliffe · · Score: 1

      Who's using 100Gbps wireless links for cellular backhaul?

    3. Re:100 gbps isn't that fast by SmaryJerry · · Score: 1

      Sprint

    4. Re:100 gbps isn't that fast by SmaryJerry · · Score: 1

      hmm sounds like the new "Lifi" that needs direct light contact

    5. Re:100 gbps isn't that fast by jratcliffe · · Score: 1

      You're off by at least two orders of magnitude. Sprint's wireless backhaul links (that they inherited from Clearwire) are 1Gbps at most - the vast majority are 100Mbps or less.

  14. But what would I be downloading? by Impy+the+Impiuos+Imp · · Score: 1

    100 Gbps is roughly 14 gigabytes per second. So I would use up my 5GB monthly Verizon cap in about 1/3 a second, then begin racking up $10 /gig overages at $140/s.

    --
    (-1: Post disagrees with my already-settled worldview) is not a valid mod option.
    1. Re:But what would I be downloading? by AbRASiON · · Score: 1

      How do you come up with those figures?
      I get a theoretical maximum of 12.5GB/sec and I've _NEVER EVER_ encountered a wireless data transmission system that comes anywhere close to it's apparent maximum. Ever.

      Wireless A,C,N,G etc - all of them, you should basically divide by 8, as standard and then halve it again. If you're lucky, maybe you'll get 60% of the claimed maximum, maybe.

      I would imagine a 100Gbps system to give a real world figure (pulling this directly from my ass as I type this,.. based on experience............) probably......... 3 to 4GB/s maybe.

  15. No distance information by Anonymous Coward · · Score: 0

    Maybe that's because its range is 1mm or less?

  16. NOT SPEED!!! by Anonymous Coward · · Score: 0

    This is a tech site, I would expect that people here understand that 'Gbps' is NOT a measure of SPEED! It is a measure of VOLUME or of course the better term that is in common use BANDWIDTH. The signal travels at the SPEED OF LIGHT!

    Continuing to mix up 'speed' & volume/bandwidth is what leads to nonsensical debates on net neutrality & rules/laws being drawn up to manage ISV's & the internet. This is especially true when making an analogy with vehicle traffic as in referring to the Internet as the 'information superhighway' & the idea that when too many people are trying to use it at the same time it is too 'slow'. The problem is NOT the speed, the problem is its VOLUME. The problem isn't that the data can't go 'fast enough' (it all travels at the same speed...either light or electrons through a wire), the problem is that there are NOT ENOUGH LANES on this highway or that the lanes aren't BIG ENOUGH (e.g. shared bandwidth).

    Please, I ask all 'proper thinking' Slashdot users to wage a campaign against using the term 'speed' in reference to bandwidth or volume of the internet.

    Now, when someone figures out how to transmit the data faster than the speed of light wake me up as I have issues with LATENCY (which is a direct consequence of the limited speed of transmission...e.g. the speed of light is too damn slow!).

    1. Re:NOT SPEED!!! by Anonymous Coward · · Score: 0

      "The signal travels at the SPEED OF LIGHT!"

      In a vacuum, maybe. Certainly not in a wire or a board.

    2. Re:NOT SPEED!!! by davec727 · · Score: 1

      Light can travel 3000 miles (across the country) in about 16 milliseconds (source: the Google) so unless you're complaining about latencies in that range, i don't think the speed of light is your problem. The packets being caught and retransmitted by many devices along the way is much more responsible for the latency.

    3. Re:NOT SPEED!!! by Anonymous Coward · · Score: 0

      Seriously? That's what you took away from my rant? You want to quibble about a few percent of 'c' in a material? The rant was about labelling bandwidth/volume measurement as 'speed'. Tech people should fundamentally know better...it's the equivalent of a physicist talking about 'speed' (magnitude only) when they mean 'velocity' (a magnitude with a direction). NO one with even a BSc in Physics would make that mistake EVER. We should not be giving in to the trope of lazy people that can't deal with 'data transmission is limited by volume'.

    4. Re:NOT SPEED!!! by Anonymous Coward · · Score: 0

      Yes...thanks very much..."I know that", but seriously, if someone can fix the 'speed of light issue' they damn well could fix the 'delay induced by multiple device issue'...having said that if someone can 'fix' the transmission of light across the country from 1 single device to another in a 'coherent manner' with only 2 devices in the way its no big deal. The best I know of today is the use of satellites as an ''intermediary' (a 3rd 'device' between 2 others) & now we're talking about significantly more than 3000 miles.

      Presumably 'switching technology' or in general 'transmission technology' will evolve & get better & thus reduce latencies in the devices used to transmit, repeat, etc. the data between a 'sender' & 'receiver'. Even today I can actually get pretty good 'ping times' from Vegas to the East Coast so I'm not REALLY complaining all that much about latency. Having said that 'bandwidth' is a 'solved problem' in the sense that it's fundamentally quite possibly simply to add more (cost is the only limitation...I'm not 'poo pooing' cost just saying that no 'new tech' is needed to add sufficient bandwidth to almost any level desired).

      The fundamentals of my rant though was 'don't confuse speed with volume'! That's akin to confusing 'speed' with 'distance', they are fundamentally different measurements.

    5. Re:NOT SPEED!!! by Bengie · · Score: 1

      The English word "speed" can be used to represent bandwidth or latency depending on the speaker. Possibly even a hybrid combination of the two.

    6. Re:NOT SPEED!!! by Bengie · · Score: 1

      Speed is just a rate at which you can do something. It takes on difference meaning depending on what the something is.

  17. 1THz = 1000GHz anybody? by Anonymous Coward · · Score: 0

    0.3THz does not make it in the Terrahertz band folk.

  18. "Close to the region used by lasers" by Anonymous Coward · · Score: 0

    Lolwut? Typical IR laser diodes used for fibre communication operate near a 1um wavelength. That translates to a carrier frequency of roughly 300THz.

    Lasers operating in the many-GHZ to THz range are called Masers ("Light" replaced by "Microwave"). The naming of the CO2 laser as a laser, and the general reference to 10um radiation as "far infrared" suggest that the desigation "laser" extends down to at least 30THz, but it certainly doesn't extend below 1THz. I'm guessing the summary is confusing the carrier frequency with the bandwidth of the modulation, which for a 10Gbps fibre is O(10GHz) per 10Gbps carrier.

    At any rate, I'm still waiting for my optical rectennas, dammit!