Multipath and Polarization

Hi Folks,

Question: Does multipath affect linear signal polarization anywhere in
the 50 MHz to 1 GHz range? For example, the low channels of US TV are
around 50 to 80 MHz, and usually the transmitters are horizontally
polarized. Would a multipath reflection, especially those encountered
in television usage (lots of concrete and steel when in a big city,
trees in the country) rotate the polarization?
-- 
%  Randy Yates                  % "Ticket to the moon, flight leaves here today 
%% Fuquay-Varina, NC            %  from Satellite 2"
%%% 919-577-9882                % 'Ticket To The Moon' 
%%%% <yates@ieee.org>           % *Time*, Electric Light Orchestra
http://home.earthlink.net/~yatescr

Randy Yates wrote:
> Hi Folks,
> 
> Question: Does multipath affect linear signal polarization anywhere in
> the 50 MHz to 1 GHz range? For example, the low channels of US TV are
> around 50 to 80 MHz, and usually the transmitters are horizontally
> polarized. Would a multipath reflection, especially those encountered
> in television usage (lots of concrete and steel when in a big city,
> trees in the country) rotate the polarization?

Would "reflection" off a "large" surface affect polarization?
I *suspect* not [based on 40 yro memory of 'first principles'].

However, "can urban environment affect polarization?" might have a 
different answer.

GOTCHA 1.
Consider a diagonal structural steel member as a diagonally polarized 
dipole. Would it not reradiate with "diagonal polarization"? Would the 
reradiated signal not be perceived to have a "vertically polarized" 
component?

GOTCHA 2.
Could not many many buildings be considered a maze of cross coupled 
leaky and lossy waveguides?

The rebar in walls and floors would define a "leaky and lossy waveguide".
HVAC ducts would be a "higher quality" waveguide.

What would the net effect of these and other "gotcha's"?

But then again, are you asking "right" question?
Back in the 60's, I installed all sorts of *low gain* *HORIZONTALLY* 
polarized antennas which solved real world problems ;)

Randy Yates wrote:
> Hi Folks,
> 
> Question: Does multipath affect linear signal polarization anywhere in
> the 50 MHz to 1 GHz range? For example, the low channels of US TV are
> around 50 to 80 MHz, and usually the transmitters are horizontally
> polarized. Would a multipath reflection, especially those encountered
> in television usage (lots of concrete and steel when in a big city,
> trees in the country) rotate the polarization?

Most structures have predominantly vertical and horizontal components. I 
don't believe they would rotate polarization much. Reflection from 
oblique planes will. Test this with a piece of Polaroid and a stack of 
glass plates.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

Randy Yates wrote:
> Hi Folks,
>
> Question: Does multipath affect linear signal polarization anywhere in
> the 50 MHz to 1 GHz range? For example, the low channels of US TV are
> around 50 to 80 MHz, and usually the transmitters are horizontally
> polarized. Would a multipath reflection, especially those encountered
> in television usage (lots of concrete and steel when in a big city,
> trees in the country) rotate the polarization?

First of all, I don't know much about EM propagation, so be warned...

I would not be surprised if polarization was changed by reflecting
surfaces. Could it be that the horizontal polarization is chosen
because it is most robust with respect to reflections? After all, the
physical ground surface is the most dominant reflector anywhere,
except in dense urban areas.

It ought to be a useful excercise both for theoreticians and
practicians to see if the polarization is changed by reflections.
Mybe this question is answered in some EM textbook? 

Rune

"Rune Allnor" <allnor@tele.ntnu.no> writes:
> Randy Yates wrote:
> >
> > Question: Does multipath affect linear signal polarization anywhere in
> > the 50 MHz to 1 GHz range? For example, the low channels of US TV are
> > around 50 to 80 MHz, and usually the transmitters are horizontally
> > polarized. Would a multipath reflection, especially those encountered
> > in television usage (lots of concrete and steel when in a big city,
> > trees in the country) rotate the polarization?
> 
> First of all, I don't know much about EM propagation, so be warned...
> 
> I would not be surprised if polarization was changed by reflecting
> surfaces. Could it be that the horizontal polarization is chosen
> because it is most robust with respect to reflections? After all, the
> physical ground surface is the most dominant reflector anywhere,
> except in dense urban areas.
> 
> It ought to be a useful excercise both for theoreticians and
> practicians to see if the polarization is changed by reflections.
> Maybe this question is answered in some EM textbook? 

Some observations by another who doesn't know much about the
subject:  A local radio and TV retransmission group orients
the FM radio antennas such that the polarization is best
for automobile antennas' reception.  If near-field polarization
effects were significant, what would be the situation in the
large antenna "farms" with numerous towers and transmitters
in close proximity?

According to an indoor measurement, there is coupleing between the two 
polarizations.  But I am not sure whether is is due to the antennas (i.e., 
purity of polarization) or the scatterers.  The trans-polarity gain of the 
channel is about 20 dB lower than that of the same polarity.  See following 
paper:

Persefoni Kyritsi, Donald C. Cox, Reinaldo A. Valenzuela, "Effect of Antenna 
Polarization on the Capacity of a Multiple Element System in an Indoor 
Environment", IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 20, NO. 
6, AUGUST 2002 page 1227.

Hope it helps.

Feng

On Sun, 05 Mar 2006 18:34:33 GMT, Randy Yates <yates@ieee.org> wrote:

>Hi Folks,
>
>Question: Does multipath affect linear signal polarization anywhere in
>the 50 MHz to 1 GHz range? For example, the low channels of US TV are
>around 50 to 80 MHz, and usually the transmitters are horizontally
>polarized. Would a multipath reflection, especially those encountered
>in television usage (lots of concrete and steel when in a big city,
>trees in the country) rotate the polarization?

I've wondered this a lot myself.   The usual Yagi outdoor TV receive
antennas are obviously set up for horizontal polarization, but that
doesn't mean that there's not some rotation in the scattered
components.

People wind up putting rabbit ears in all sorts of odd arrangements,
often vertical, and I always took that as a hint that once it got
indoors the scattering could be almost anything.

Above 1GHz it's a bit weird as well.  Professional installations of
WLAN APs almost always orient the antennas orthogonally.   Laptops and
PDAs can be in any old position and are still expected to work, so
either most antennas don't care about polarization (is that possible?)
or scattering is expected to be rich enough that it doesn't matter.

I'm kinda pissed at our antenna guy right now or I'd ask him for you.

Eric Jacobsen
Minister of Algorithms, Intel Corp.
My opinions may not be Intel's opinions.
http://www.ericjacobsen.org

"Randy Yates" <yates@ieee.org> wrote in message 
news:m3slpw4u4z.fsf@ieee.org...
> Hi Folks,
>
> Question: Does multipath affect linear signal polarization anywhere in
> the 50 MHz to 1 GHz range? For example, the low channels of US TV are
> around 50 to 80 MHz, and usually the transmitters are horizontally
> polarized. Would a multipath reflection, especially those encountered
> in television usage (lots of concrete and steel when in a big city,
> trees in the country) rotate the polarization?
> -- 
> %  Randy Yates                  % "Ticket to the moon, flight leaves here 
> today
> %% Fuquay-Varina, NC            %  from Satellite 2"
> %%% 919-577-9882                % 'Ticket To The Moon'
> %%%% <yates@ieee.org>           % *Time*, Electric Light Orchestra
> http://home.earthlink.net/~yatescr


Hello Randy,

In a nutshell, the answer is yes - polarization gets messed with. In a 
single path environment, the polarization may be either rotated by 
reflection off of an obliquely oriented non-metallic object or reflection 
off of a metallic object. In the 1st case, the reflection is still linearly 
polarized. The 2nd case is more complicated. The plane parallel and plane 
perpendicular components will each reflect with a different reflection 
coefficients and phase shifts. So in general, metallic reflection creates 
elliptical polarization. This happens all across the EM spectrum including 
radio and light waves. For the nonmetallic case, the phase shifts are the 
same although the reflectivies are still different - so linear stays linear.

You probably recall from your Physics studies about the index of refraction 
which simply is the ratio of the speed of light in a vacuum to the speed of 
light in matter. Also you will recall that this usually is frequency 
dependent, hence this is why a prism decomposes white light into a 
spectrum - each frequency has a different velocity in the glass and because 
of the oblique incidence is refracted by a different amount.

If one applies Maxwell's equations and the continuity relations to the 
problem of wave propagation across a matter boundary, one arrives at 
Fresnel's equations for the reflection and transmission coefs based on the 
indices of refraction. It turns out that metals can be properly described as 
having complex indices of refraction where the imaginary part corresponds to 
the absorptivity of the material. For example, aluminum has an index of 
refraction of 1+4.45i at the wavelength of 632.8 nm (the orange-red line 
produced by a He-Ne laser.) And from Fresnel's equation one finds that 
linearly polarized light when incident at 78 degrees and oriented so both 
the plane parallel and plane perpendicular components have equal power upon 
incidence, you will create circularly polarized light. Back when I used to 
make holograms, I showed how to match the polarization rotation in each of 
the light beams so the light's interference was maximized. A wrote a CAD 
program to track the effects of metallic polarization caused by the aluminum 
mirrors used in the optical setup. I was able to design polarization 
independent optical setups where a less expensive non polarized laser may be 
used.  A science based on metallic polarization used to measure thickness of 
thin films is called ellipsometry.

Now back to the RF. Since any arbitrary polarization can be decomposed into 
the sum of other polarizations, you can now see the a multipath situation 
involving metallic objects can result in some pretty screwed up 
polarizations. But even without the metal components, the polarization can 
get rotated around if your received signal ends up being composed of mostly 
reflected energy. Even with vertical reflecting structures, if you are near 
the transmitting antenna and its height is different from your receiving 
antenna's height, you have a certain amount of obliqueness, and can see a 
rotated polarization.

IHTH,

Clay

"Randy Yates" <yates@ieee.org> wrote in message 
news:m3slpw4u4z.fsf@ieee.org...
> Question: Does multipath affect linear signal polarization anywhere in
> the 50 MHz to 1 GHz range? For example, the low channels of US TV are
> around 50 to 80 MHz, and usually the transmitters are horizontally
> polarized. Would a multipath reflection, especially those encountered
> in television usage (lots of concrete and steel when in a big city,
> trees in the country) rotate the polarization?

There is a rule of thumb that suggests that horizontal polarisation is 
marginally better than vertical, and so that is predominantly used for fixed 
point comms in VHF/UHF. The school of thought here is that over a longer 
distances, the effects of fading due to Brewster angle cancellation is 
minimised, and closer in that 'most' ground based reflecting (multipath 
creating) objects are vertical (posts, pylons, etc). I have also seen 
anecdotal reference to horizontal polarition being less likely to switch to 
vertical rather than vice-versa over terrestrial links.

For mobile links, vertical polarisation is preferred solely because it's a 
lot easier to manufacture an ergonomic antenna producing an omnidirectional 
pattern from a vehicle or handset with a vertical antenna than with a 
horizontal one - a loop mounted 3' above a car looks horrible compared to a 
vertical whip.

For VHF broadcast band II transmissions (ie those on the FM dial), slant 
polarisation is often used at the transmitter to cater for both fixed and 
mobile receivers.

In Europe, in general primary UHF TV transmitters are horizontal and the 
fill-in spots are vertically polarised.

In general, as another poster mentioned, because our buildings are made of 
horizontal and vertical edges, polarisation angle change due to reflection, 
diffraction (and to a lesser extent refraction when travelling through 
material) happens but the degree of polarization change is somewhat limited.

If you are concerned about the whether two terrestrial signals in V/UHF 
sharing the same spectrum but at different polarisations will interfere with 
each other, then expect it. On the other hand, do not be dependent on the 
phenomenon either! This is where techniques like adaptive antenna diversity 
come into their own.

On the other hand, for line of sight microwave links, like Ku band direct 
broadcast satellite TV, there are no significant polarisation changes to 
talk of, and so spectrum reuse using the two orthogonal polarisations is 
used. The same cannot be said the V/UHF where the phenomenon of Faraday 
rotation kicks in - this is why circular polarisation is used for V/UHF 
space comms, but that's a whole new story.

Of course if we lived in Egyptian times, our choice of polarisation might be 
somewhat different.

Regards, Howard

Hi Clay,

Wow! Lots of information here. Thanks for sparking my thoughts. 

> [...]
> So in general, metallic reflection creates elliptical polarization. 

Elliptical polarization? I'm not getting it. I thought the EM wave has the
E field orthogonal to the B field, and both of these are orthogonal to the
direction of travel. So where does an ellipse come in? Or do you just mean
something between horizontal and vertical (e.g., 78 degrees)?

> IHTH,

Yes, it does. Thank you, Clay.
-- 
%  Randy Yates                  % "Though you ride on the wheels of tomorrow,
%% Fuquay-Varina, NC            %  you still wander the fields of your
%%% 919-577-9882                %  sorrow."
%%%% <yates@ieee.org>           % '21st Century Man', *Time*, ELO
http://home.earthlink.net/~yatescr