: Scientists Discover How To Boost Wireless Bandwidth

2012-03-02, 09:18 AM
Here is a very interesting story about how scientists in Sweden and Italy have successfully tested a hypothesis that OTA bandwidth capacity can be significantly boosted by adding a "twist" to the transmission signal.

'Twisted' waves could boost capacity of wi-fi and TV:

2012-03-02, 06:17 PM
Heh, yeah I saw that this morning too. But the article I saw didnt really explain the difference between it and circular polarization, which has been around for a long time. I guess its the spacing of the "twists" that makes the difference.

2012-03-02, 07:24 PM
It does sound a little like polarization but vertical + horizontal, not circular. That has been done with satellite signals for years. The fact that they demonstrated only double the bandwidth makes me question the science behind the demonstration.

2012-03-02, 08:32 PM
Its not increasing the bandwidth per se, but rather more like putting 2 lbs of crap in a 1 lb capacity bag, heh.

2012-03-03, 12:11 AM
I saw this posted else where by a fellow named Ai4i.
I've been trying to understand this for a while now, he said.

A friend lived 200 Km from a CP TV station he enjoyed viewing, but an offending horizontal station was another 150 Km in the same direction, and whenever it got cold, the distant station would cause objectionable QRM. Using a vertical antenna would not resolve the issue because extended propagation always randomizes polarity. He built two CP receiving antennŠ, one CW and one CCW and connected them out of phase. They both received the linear station and their combined output was next to zero, but only one of them received the target station. It seemed counterintuitive, but connecting and disconnecting the second antenna had no effect on the strength of the desired station, but connecting it make the offending station go away

The reason I posted it here is how similar this is (two antennas connected out of phase ) to what the researchers did described in more detail here (http://iopscience.iop.org/1367-2630/14/3/033001/article) "Bo ThidÚ" is mentioned in both so I'm pretty sure it's the same experiment.

The signal was collected equally by antennae A and B in phase and the signal of antenna A arrived at the signal adder 180░ out of phase with respect to that of antenna B because of the electric λ/2 cable delay, resulting in a difference signal configuration

Looks a lot like what I'm still not sure what they did new.
Wifi enthusiasts and others have known a long time that twice as many users can use the same frequencies if half use vertical polarized antennas and half use horizontal.

And I thought part off how N routers with "mimo" got more band width from a single channel was by decoding separate streams based on phase/time differences.

Scarry bob said
It does sound a little like polarization but vertical + horizontal, not circular actually in the link(above) it looks like one was linear and one was circular( but then they seem to contradict themselves). I to share your skepticism of actual throughput improvement. And wonder if the "test " could have been conducted with any two oppositely polarized antennas (linear or circular)

I also noticed they tested over water, think a ridge detraction might hurt the effect ?
Still wading.

2012-03-03, 12:04 PM
Wifi enthusiasts and others have known a long time that twice as many users can use the same frequencies if half use vertical polarized antennas and half use horizontal.
And I thought part off how N routers with "mimo" got more band width from a single channel was by decoding separate streams based on phase/time differences.

Well, the multi-user data through-put capacity of Ethernet, whether on Cat-6 cable or a single wifi channel, is limited by the number of collisions. More collisions, less net data through-put. Anything that minimizes collisions will increase the multi-user data through-put.

2012-03-03, 12:12 PM
actually in the link(above) it looks like one was linear and one was circular
I also forgot that circular polarization can be reversed. Either way, I'm not yet convinced this "discovery" is anything new.

2012-03-03, 12:17 PM
Either way, I'm not yet convinced this "discovery" is anything new.
Yeah, and I dont think this "discovery" could possibly work in practice for long distance DTV signals with only small changes in existing equipment. :p

2012-03-03, 03:15 PM
whether on Cat-6 cable or a single wifi channel, is limited by the number of collisions

There no collisions on Ethernet these days, with switches and full duplex commonly used. Collisions occured on the old coax and hub based networks. With switches, there is no longer a collision domain to worry about.

I also forgot that circular polarization can be reversed

I also get the impression they're talking about circular polarization.

2012-03-03, 04:23 PM
Jamesk is right on the ethernet but 802.11 abg and I think N use (CSMA/CA) to deal with collisions and all devices within ~3 channels share a collision domain.
Just to clarify my statement about Wifi, the devices can tune to twelve channels +or - depending on your country but only 3 are far enough apart to not interfere, without large attenuation channel 3 interferes with
1 - 6 and 9 with 6 -12. But with the attenuation provided by opposite polarization channels 3 and 9 can be used like this ch. 1-VP. ch. 3-HP. 6-VP. 9-HP 12-VP effectively adding 2 more usable channels/wires. This works good outdoors with LOS but not so much inside buildings where multiple reflections can totally jumble the polarization.
I think that's what's happening when you rotate your biquad and see no drop in signal it's because the signal got jumbled (polarization wise) before it finds it's way to the window and out.
Theoretically 1, 6. and 12 could all have two transmitters on each, one HP and one VP and not even see each other. (but those are spherical chickens in a vacuum)

Meanwhile back on the subject : I have to agree that this technique is unlikely to help long distance tv. The refraction of 1 edge likely kills the
signal integrity beyond recovery. And it's hard to imagine a tropo signal that would tolerate a big null (singularity ) at the boresite.
They picked over water for a reason.
IMHO It is more likely to kill DX reception if used for tv transmission hopefully it will be relegated to LOS data links and microwave relay stations ect. not TV.

I do think they have discovered something, remember someone "discovered " America and electricity and in both cases it wasn't the first person to see or notice it !

I guess the split does something but not polarization (diffraction?) because it says both signals were linear, My mistake. I should alter a parabola like they did to see if what nec shows going on
(if it shows anything). I don't think the split causes circular polarity (I thought the feed was CP).

2012-03-04, 07:51 AM
I saw the story too. Am I correct in thinking that everybody would need to buy new receivers if broadcasters were to start to use this ?

2012-03-04, 08:00 AM
Given how vague the description is, I expect we'll have to see something far more concrete before we have to worry about anyone switching.

2012-03-04, 11:33 AM
I managed to finish the article (that doesn't mean I caught it all ) and I picked up a few facts.

Am I correct in thinking that everybody would need to buy new receivers if broadcasters were to start to use this ?

No they use all "COTS" equipment that's "consumer off the shelf" (except the split dish)

It's all in the "interferometer" (two yagis connected 180 degrees out of phase) and the separation distance between and aiming of the two yagis determines what signal is left for the tuner to pick up.

With one yagi you would have heard both tones/signals.

Already with this setup, one can obtain four physically distinct channels on the same frequency by additionally introducing the use of polarization (SAM), which is independent of OAM. A further five-fold multiplicative factor from implementing multiplexing would yield a total of 20 channels on the same frequency. The utilization of multiport techniques (e.g. MIMO) could increase the capacity further.
So I was wrong it's not the same as either polarization or mimo.

This PDF is the same report I think.


I'm still not sure what the exact effect of the split dish but if (that's a big if)edge diffraction imparts similar qualities in the signal then this could be used to not just attenuate offending co-channel interference but pick which channel to receive (if both signals were strong enough).

2012-03-04, 12:08 PM
Thanks for the link. :p

Helicoidal parabolic antenna is what he calls his new antenna.


It does seem like it may be good for tower to tower LOS applications. But in practical terms, I think it would mean a separate antenna for each channel on the same frequency, which would mean the number of channels would be limited by the size of the tower ??

And there would be no such thing as a broadband helicoidal parabolic antenna ?? (but multi channel maybe ?)

As usual, the solution for one problem may cause a host of other problems, heh.

2012-03-05, 05:45 PM
OAM (Orbital Angular Momentum), aka "Radio Vorticity" aka "Twisted Radio Beams":

OAM is a fairly well-known capability for Lasers and has been observed in Space.
But Tamburini's 1Mar2010 article is the FIRST report of a public On-Air Test using
Radio signals (2.4 GHz Wi-Fi Band).

Generating a OAM signal can be accomplished with a circular array of antennas, each
fed with the same signal, but adjacent antennas have a phase difference = 2*Pi*l/n,
where n=number of antennas and l=integer (e.g. 1, 2, 3, 4, 5, etc), so that the
phase rotation around the entire circular array is 2*Pi*l.

The Venice Experiment injected OAM by using a Horizontally Polarized Parabolic
Antenna cut from the outer edge to the center and separated by 6.25-cm at the outer
rim, thereby generating the increasing phase difference for l=1. Yagi's would be l=0.
They only built one Parabolic to generate l=1. If they had used two additional
Parabolic Antennas with the Gap twice and three times as big, they could have
transmitted THREE orthogonal OAM signals on the same frequency, thereby
TRIPLING total capacity.

At the receiver, two identical, Horizontally Polarized Yagi Antennas were summed
together, except one of them went through an inversion, so the output was the
DIFFERENCE SIGNAL. When receiving from OAM Parabolic, the output of this
"Interferometer" was the OAM Signal. A parallel, Yagi Antenna link was
transmitting at the same time on the same frequency. Transmit power was
adjusted so that the received signal level matched the interferometer output,
thereby adjusting for the much higher Gain in the Parabolic than the Yagi and
overcoming whatever losses were in the Interferometer. I THINK that the
interferometer had to be reconfigured to switch between the Yagi signal (l=0)
and the Helicoidal Parabolic signal (l=1)....or maybe they used another Yagi
at the receiver????

They had to carefully move the receive interferometer around until they found the fairly
small "sweet spot" where the OAM signal could be detected.

I would like to see how much "leverage" the OAM signal has versus conventional,
since it appears to be a minor "tweak" to the primary signal, so I would expect the
sensitivity would be much less....and how much loss was there in the interferometer....

And I don't know what they had in mind when they claimed an additional 5X increase
in capacity using "multiplexing" techniques.....

Current OAM antennas are clearly limited to point-to-point links, vice broadcast.

I followed several references....below are a "must read".

Fabrizio Tamburini et.al. - Encoding Many Channels in Same Freq thru Radio Vorticity -
First Experimental Test - 1107.2348 - 17pgs 1Mar2012.pdf
http://arxiv.org/ftp/arxiv/papers/1107/1107.2348.pdf [Better Quality Graphics]

Bo Thide + Fibrizio Tamburini et.al. - Radio Beam Vorticity and Orbital Angular Momentum -
1101.6015 - 3pgs 31Jan2011.pdf
http://arxiv.org/pdf/1101.6015.pdf [A bit more info re Venice Experiment.]

H Then + B Thide et.al. - Detecting Orbital Angular Momentum in Radio Signals - 0805.2735 -
5pgs 18May2008.pdf [THEORY, Tripole & Multi-Element Arrays]

B Thide et.al. - Utilization of Photon Orbital Angular Momentum in Low-Freq Radio Domain -
0705.1208 - 4pgs 1May2009.pdf
http://arxiv.org/pdf/0705.1208.pdf [Multi-Element Antenna Array]

Bo Thide - Nonlinear Physics of the Ionosphere and LOIS/LOFAR - 0707.4506 - 6pgs 31Jul2007.pdf
http://arxiv.org/pdf/0707.4506.pdf [Multi-Element Antenna Array]

These were conveniently found via Columbia Univ's Archive Search Engine:

arXiv.org Search for Downloads - Tamburini_F

arXiv.org Search for Downloads - Thide_B

Some additional references:

R. Compton, Jr - "The tripole antenna: An adaptive array with full polarization flexibility"
[IEEE is a subscription service....or visit a well equipped Univ. Library]
http://ieeexplore.ieee.org/Xplore/login.jsp?url=http%3A%2F%2Fieeexplore.ieee.org%2Fiel6%2F8%2F 25617%2F01142690.pdf%3Farnumber%3D1142690&authDecision=-203

H.T. Hayvaci et.al., Univ. of Chicago, Presentation: Tripole antennas for Wi-Fi:

2012-03-06, 01:01 PM
Orbital Angular Momentum (OAM) can be independently transmitted on BOTH
Horizontal & Vertical Polarization (or Right Hand & Left Hand Circular Polarization),
thereby increasing the total number of simultaneous signals by DOUBLE....
or TRIPLE the capacity using a Tripole Antenna...but NOT along the NULLS:
Note this article only addressed Linear Polarization, i.e. Spin Angular Momentum (SAM).

A minimalist OAM TX/RX System is described as a planar array of 4 (or 3?) antennas,
each capable of sensing the 3D nature of the E&M field, such as a Tripole or a
Cubical Antenna. Since each Tripole would require THREE Receivers, the total
number of receivers would be either 12 (or 9?). For the Cubical Antenna, there
are twice as many sensor locations and hence a minimum of 24 (or 18?) Receivers.

Although my D-Link Wi-Fi Router has THREE antennas...and hence three receivers,
building an OAM receive/antenna system would be fairly expensive....plus what
appears to be direct RF (or IF?) Sampling (with ZERO phase error) feeding a
much higher speed signal processor(s).

It is important to observe that the Venice Experiment used an Interferometer
with a SINGLE Receiver....and very careful alignment of a pair of Yagi Antennas....
So a simpler RECEIVE antenna may be possible when not trying to image a Space object.

PS: An orthogonal Triple Loop antenna could also be used:

Some additional references of interest:

Bo Thide at Boulder CO - OAM in Radio Experiments - AN OVERVIEW - 59VUS 2009.pdf

J Sjoholm + K Palmer, MSc Thesis - Angular Momentum of E-M Radiation - 196pgs Apr2007.pdf
All the gory theoretical details....or skip to the interesting parts re OAM....

Bo Thide - Tripole Arrays + OAM - LOFAR and LOIS Radio Telescope - 53 VUs 2005.pdf

Sandra Eriksson, MSc Thesis - Study of Tripole Antenna Arrays for Space Radio Research - 95pgs 2Sep2003.pdf
Excellent Analysis. Also prepared under Prof. Bo Thide's guidance (Uppsala Univ, Sweden). Includes
phase shifts between antennas, but was before OAM techniques were being explored in radio systems.

T Hult + A Mohammed re Tripole + Triple Loop - MIMO Antenna Apps for LEO Sat Comm -
MIMOWS3-022 - 16VUs .pdf

There are very few FREE articles re "Tripole Antenna". You can find articles
for this and other subjects via IEEE Explore Search Engine [NOT FREE], such as:
Kajfez, D. + Harrison, M., U of Mississippi, "Experimental Investigation of the Tripole Antenna"

Mohammadi et. al, "Orbital Angular Momentum in Radio—A System Study"

BTW: "Tripole" also describes a Folded Dipole (or Director) with an extra Element, which
I call a "Tri-Fold" to avoid confusing the two entirely different antenna structures:

Some additional Publications & Thesis Documents from Swedish Inst. of Space Physics:

2012-03-06, 09:22 PM
OAM Clarification:
In post #15, the eigenvalues can ALSO be negative, e.g. l = -1, -2, -3, -4, -5 etc.
A negative eigenvalue would correspond to a Parabolic dish cut from rim to center,
except with the OTHER side of the cut being brought forward.

The above cited: J Sjoholm + K Palmer, MSc Thesis - Angular Momentum of E-M Radiation
explains that increasing numbers of antennas in the array are needed to support higher
eigenvalue numbers. For example, 10 Antennas were needed to support l=1-4. So if
Tripole Antennas were used, 30 Receiver Systems would be required to QUADRUPLE
the total transmit capacity due to OAM and DOUBLE again due to Polarization...
And EIGHT Antenna/Transmitter Systems. Note the ever increasing cost penalty.....