back to article US boffins demo 'twisted radio' mux

Evidence continues to mount that the orbital angular momentum (OAM) of radio waves – “twisted waves” – can be exploited to modulate multiple data streams on the same spatial path. Researchers from the University of Southern California are claiming an impressive 32 Gbps transmission, albeit over a distance of 2.5 metres, using …

  1. frank ly

    Numbers always confuse me

    " .. a spectral efficiency of 16 bits/second/Hz on a 28 GHz carrier." gives 448 Gbps ?

    "claiming an impressive 32 Gbps transmission, ........, using eight OAM channels – four independent OAM beams on each of two polarisations."

    Can anyone explain that?

    1. Anonymous Coward
      Anonymous Coward

      Re: Numbers always confuse me

      From what I can make out, a single OAM beam can do up to 4Gbps. Sending two beams at the same frequency but at opposite polarisations is sort of like double-pumping: it's possible because the two beams are out of phase to each other so don't cross-talk. Now use four frequency bands and Bob's your uncle.

      That said, I have an honest question of my own. These beams seem very directional to me. Wouldn't this mean that the receiving end has to be aligned just right to correctly receive the beam?

      1. Paul Kinsler

        Re: the receiving end has to be aligned

        Yes; hence the "along the same spatial path" remark. The axis of the beam would have to hit inside the outline of the receiver so that the different OAM could be distinguished.

      2. JeffyPoooh
        Pint

        Re: Numbers always confuse me

        "Sending two beams at the same frequency but at opposite polarisations is..."

        ...old hat.

        That's how satellite TV works. FSS uses H/V and DSS uses RHCP/LHCP. It's bordering on ancient.

        I'm not sure if they're talking about the same thing here. They can't be that daft to not be aware of the above factoid. I'm not yet sure what the hell OAM actually is, but it has to be something different than just polarization.

        1. Paul Kinsler

          Re: I'm not yet sure what the hell OAM actually is ...

          Ahem:

          http://en.wikipedia.org/wiki/Orbital_angular_momentum_of_light

          1. JeffyPoooh
            Pint

            Re: I'm not yet sure what the hell OAM actually is ...

            Yeah, of course I read the wiki and a few more. Still hasn't sunk in.

            If the boffins would implement it at UHF and draw a schematic of the antenna system, maybe it would sink in.

    2. John Robson Silver badge

      Re: Numbers always confuse me

      Not using spectrum from 0Hz to 28GHz, but some bandwidth around the 28GHz carrier...I think

      1. frank ly
        Facepalm

        @John Robson Re: Numbers always confuse me

        Of course!

  2. Christian Berger

    I'd still say it's MIMO

    Just because you combine multiple antennas into one, doesn't mean you can't have different channels.

    1. Charles 9

      Re: I'd still say it's MIMO

      I thought the key element to MIMO is the fact it uses multiple antennae in order to take advantage of interferometry to improve signal clarity. IOW, this can't be MIMO as most would understand the concept.

      1. Anonymous Coward
        Anonymous Coward

        Re: I'd still say it's MIMO

        I have a crossed beam antenna - essentially 2 yagis at 90º on the same boom; spaced so one is slightly in front of the other.

        Either polarization is usable independently, or with an appropriate phasing harness (carefully sized bits of coax leading to each yagi, connected to one transmitter) I can use both at the same time to produce circular polarisation.

        Does that count as MIMO?

        1. JeffyPoooh
          Pint

          Re: I'd still say it's MIMO

          "Does [my Crossed Yagi] count as MIMO?"

          No.

          A good example of MIMO are the wifi routers with three antennas. Usually.

          1. Christian Berger

            Re: I'd still say it's MIMO

            Actually crosses Yagis are, if they have independent outputs, MIMO capable. In fact MIMO for LTE uses just such constructions. You have multiple antennas or antennas which have 2 outputs one for each polarisation.

            MIMO just means Multiple Input Multiple Output. It has little to do with what sort of antenna you use.

            1. JeffyPoooh
              Pint

              Re: I'd still say it's MIMO

              *His* doesn't, and isn't. If it were redesigned, perhaps it could be (but same as a coat hanger could be). MIMO is more than just an antenna. MIMO has to include a box of tricks in the router itself. And every one that I've seen has at least three antennas.

              So multiply redundant No.

              I'm using "MIMO" to mean "MIMO", not just the words.

  3. Paul Crawford Silver badge

    Sceptical

    AFIK the idea of OAM is the polarisation is rotating. Now you can generate any polarisation by taking a pair of orthogonal antennas and driving them with the appropriate amplitude & phase

    You get linear at any angle if the phase shift is zero, with the angle determined by the magnitude of the two drives.

    You get circular with LHCP or RHCP depending on the phase being +/-90 deg.

    So if you were to drive the amplitude in a cyclic manner you would get the appearance of a rotating linear phase, and if at the receiving end you were to combine the similar antennas with a matching cyclic ratio then bingo - you have the original signal as if it were received by a rotating antenna.

    But how is that different from any classical modulation on dual polarisations? Sure they might be claiming the equivalent of higher than QPSK-like "polarisation constellation" points, but that is not without a loss of orthogonality and hence some cross-talk and loss of SNR.

    The real question then is can such a scheme deliver any better then just going to higher RF modulation constellations on two classical orthogonal polarisations?

    1. Paul Kinsler

      Re: Sceptical

      An ordinary beam with rotating polarization has no interesting transverse behaviour, and in most systems whatever the spatial profile is, it is ignored. So with both polarizations you get two channels not one.

      These OAM beams have a nontrivial transverse spatial modulation; and that spatial structure can carry extra information, if the receiver can detect it.

      1. JeffyPoooh
        Pint

        Re: Sceptical

        One thing to keep in mind is that CP received by linear, or vice versa, are compatible, but there's a -3dB loss due to the polarization mismatch. This 3dB eats into Shannon's window. I wonder if they've opened up another channel that carries precisely the same amount of information as is lost in the -3dB SNR reduction?

    2. Anonymous Coward
      Anonymous Coward

      Re: Sceptical

      I'm with Paul. Consider satellite: In the US, DBS from Directv uses the same frequency range (11700-12200 MHz for Ku; 17300-17700 / 18300-18800 / 19700-20200 MHz for RDBS / Ka lo / Ki hi) for simultaneous LHCP and RHCP polarized signals, which use QPSK or in some cases 8PSK modulation.

      So the question is, would what is described in the article allow them to utilize the same amount of spectrum more efficiency than they already are? i.e. multiple LCHP beams with different "twisted" characteristics? Or is that going to increase the noise floor by enough that instead of QPSK/8PSK they have use a lower order modulation or increased error correction so there is little or no increase in actual usable bit rate?

      1. JeffyPoooh
        Pint

        Re: Sceptical

        Is this our emerging consensus? That the increased noise floor introduced into the legacy channels will eat into their informational bandwidth in precise balance to the bandwidth offered by the new channel.

        It was about ten years ago I imagined a techique to hide some linear signals amongst the RHCP and LHCP, in such a way that nobody would notice. After thinking about it a bit longer I realized that the hidden signals would still increase the noise floor of the CP signals, so it really didn't increase the total data rate. It seemed to be a zero sum game.

        Ideas - a dime a dozen.

        1. Anonymous Coward
          Anonymous Coward

          Re: Sceptical

          Don't mistake my question for consensus. I don't know nearly enough about RF to know whether that's the case. I don't think that those working on this "twisted" stuff are pursuing a dead end, but when it is translated through journalists who don't know science that well they may not understand the limitations. Maybe it only works in certain circumstances (certain frequency ranges, certain power ranges, certain antenna sizes, etc.) that maybe make it useful for mobile devices but not satellite, or satellite but not UHF broadcast, or UHF broadcast. Maybe it works in the air but not over a wire for more than a few cm.

          Lots of possible roadblocks to the hype portrayed in press reports about this, but I'm not qualified to tell what's what :)

  4. Destroy All Monsters Silver badge
    Coat

    That image

    So Hitler was on to something?

  5. Anonymous Coward
    Anonymous Coward

    Spiral phase plate antenna?

    I'm no expert but there is no way this "spiral phase plate" antenna can impart 'twisted waves' to an RF transmission. This is the data propagation version of a perpetual motion machine. Besides which, there's no way of extracting a modulated signal from the orbital angular momentum of a RF signal.

  6. bminish

    I fail to see how this is any different from using Horizontal and Vertical Polarization. H/V if carefully set up allows for 2 separate paths with good Isolation. This is Widely used today

    1. JeffyPoooh
      Pint

      I fail to see...

      I don't understand it (yet) either.

      Perhaps they're using the polarization itself as another carrier. I can't imagine this would do more than slightly open up the Shannon window for additional bits. I'd be shocked if if were a doubling of throughput.

      Problem is, these academics are so hungry for fame that they issue press releases for the most mundane and/or idiotic things sometimes. One can't tell if it's a real thing or another BS publicity stunt. It's a wonder they didn't mention that "the rotating wave are 3D printed."

      1. Hargrove

        Re: I fail to see...

        I'm an old microwave component engineer who spent the last 45 years as a systems engineer. In other words, I have lost most of what I ever knew about the subject. But, from what little I do remember, I found the quality of the discussion is impressive.

        @JeffyPoooh is right on the mark.

        I'm reminded of a conference I attended where a professor from a major US university truncated his presentation to introduce a revolutionary new invention for eliminating household electrical wiring that his graduate students had discovered. This new discovery had the astonishing ability of being able to transmit electrical power through the standard construction gypsum wallboard interior walls.

        He then showed a picture of the prototype, which consisted of two large coils on either side of the wall, one driven by the power source, the other hooked to a light bulb. No one in an audience of about 400 people had the temerity to point out the obvious. . . that his geniuses had re-invented the resonant air core transformer. I would have, but I was too busy wiping up the coffee that I had just spurted out my nose.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like