I am a little skeptical of path analysis software that attempts to calculate 
multipath in an outdoor environment at the frequencies and over the distances 
that WISPs use.  Just some back-of-the-envelope calculations:

 

Take the example of a 2 mile path across flat ground between antennas each 20 
feet AGL, with a reflection off a pond in the middle of the path.

 

Additional distance for multipath reflection = approx. 1 inch

Wavelength at 5 GHz = approx. 2 inches

 

So in this example (which ignores earth curvature), the reflected path is right 
around a half wavelength longer.  If the path is longer or the antennas are 
higher, the difference will be several wavelengths.  I think this is why we see 
multipath problems change rapidly with crop growth, ground fog, thermal 
inversions, and even changes in air temperature.  At lower frequencies like 
AM/FM/VHF/TV or on shorter paths (like indoors), it would not be as critical.

 

I know that theoretically OFDM is more resistant to this because each 
subcarrier sends symbols that stretch out over a much longer time period than 
1/f.  But it seems to me that a software calculation based on topographic and 
clutter databases cannot be accurate down to an inch or even a foot, plus other 
factors like air temperature and vegetation growth will introduce more than an 
inch of variation.

 

The other thing is that I’ve never seen this supposed OFDM immunity to 
multipath actually work in the field.  For example, when I swapped out hundreds 
of Trango single carrier radios for Cambium 430 and 450, the customers with 
multipath problems had exactly the same multipath problems afterward.  Same 
when we swapped out FSK for 450.  And remember how we were supposed to use 1/8 
or 1/4 cyclic prefix on 430 systems to maximize immunity to multipath?  I never 
found it to make a difference, other than reducing throughput.  Evidently 
nobody else did either, because now the only choice is 1/16 CP.

 

This is not a rigorous analysis and may very well be wrong, but unless someone 
can cite field experience to the contrary, I’m a skeptic.  Not talking about 
indoor bouncing off walls and furniture, that may very well work.

 

 

From: [email protected] [mailto:[email protected]] 
On Behalf Of Lewis Bergman
Sent: Wednesday, December 21, 2016 8:04 AM
To: Cambium Networks User Group <[email protected]>
Subject: Re: [Cambium-users] Multipath

 

Protocols like OFDM can use multi-path (time diverse bit reception) and piece 
those back together to increase signal strength and SNR. Most if not all the 
currently offered products including ePMP and 450(i)(m) as well as most of the 
PTP products use MIMO OFDM or Massive MIMO OFDM which are capable of piecing 
together bits arriving at different times to enable a clearer constellation and 
a higher bit rate.

 

On Wed, Dec 21, 2016 at 7:25 AM Sam Morris <[email protected] 
<mailto:[email protected]> > wrote:

When creating propagation maps in TowerCoverage there is a feature
called "Two Rays". The explanation in the TowerCoverage WIKI says "Two
rays are used for multi-path calculations. Most WISP use radios that
cannot do multi-path reflections so by default this option is turned
off. If your radio system can use multi-path reflections then you may
want to turn this value on."

In this context, do Cambium radios use multipath reflections? Whenever
I've asked about this before, it seems I was told to leave this off
unless you're shot goes over a large body of water.

_______________________________________________
Cambium-users mailing list
[email protected] <mailto:[email protected]> 
http://lists.wispa.org/mailman/listinfo/cambium-users

_______________________________________________
Cambium-users mailing list
[email protected]
http://lists.wispa.org/mailman/listinfo/cambium-users

Reply via email to