[TowerTalk] Diffraction Reciprocity

[David Gilbert]

I asked myself that question a dozen times up front, and always got the 
answer that since diffraction is a function of the radius of curvature 
of the diffracting surface, if the diffracting surface is asymmetric 
(different radius of curvature for the incident wave as it travels in 
either direction) the effect of the diffraction cannot be completely 
reciprocal.  Whether it is valid to say that diffraction itself is 
non-reciprocal is a semantics issue on which I'm no expert.

An incident wave does not act like an infinitely thin ray .. it acts 
like a wave, and it is influenced by its surroundings within a distance 
determined by its wavelength.  A wave that passes close by (in terms of 
wavelength) a surface will suffer dispersion and loss that is a function 
of the radius of curvature of the surface ... an edge with a small 
radius disperses the wave (some of which bends around the surface) more 
so than a surface with a larger radius.  You can either accept that or 
not ... it isn't my job to convince you.

My original postulation was that since terrain features are sometimes 
not symmetric, it seemed that an outgoing wave might be diffracted 
differently than a return wave attempting to follow the same path.  I 
later discovered that N6BV, the author of HFTA, had said pretty much 
that same thing in the ARRL Antenna Book.  If I get the chance I'd like 
to try to demonstrate that experimentally ... glad to have your blessing.

A few additional comments.

First, the favorable bending (diffraction) of a signal over a terrain 
feature is mostly noticeable at very low angles.  If the path between 
two stations is such that low angles aren't required, or propagation is 
such that low angles are useful but higher angles are significantly 
involved as well, then whatever effect that may be caused by asymmetric 
terrain is not likely to be noticed.  I think the very great majority of 
actual QSOs would not be affected.

Secondly, the actual elevation profile of the signals coming from either 
direction is not a simple function of diffraction.  It's a function of 
the combination of the reflections and diffractions that occur because 
of the terrain.  So far in this discussion I've only mentioned 
diffraction because it seemed to me that we if established the 
asymmetric effect of diffraction on the path that would be sufficient to 
suggest a mechanism for non-reciprocal transmission of signals.  But in 
reality, it is almost impossible to assume that all those combinations 
of reflection and diffraction can occur identically from either side of 
an asymmetric terrain feature, making the case (in my opinion) even 
stronger.

Thirdly, a wave leaving Station A with horizontal polarization will 
interact with its foreground in a manner calculated by HFTA.  
Conventional knowledge says that after passing through the ionosphere 
the polarization of a signal becomes mostly elliptical.  Since 
diffraction (I think) and reflection (I know) are a function of 
polarization, it would seem that the reverse direction would behave 
somewhat different based upon that alone, although probably not by very 
many db.  Still,  it might be relevant when signals are in the mud.

Finally, I'm not searching for all the possible mechanisms that might 
possibly create non-reciprocal signal readability.  I'm not even trying 
to prove that non-reciprocal communication exists ... only that IF it 
exists, asymmetric terrain at one end of the path may be a possible cause.

73,
Dave   AB7E




Ethan Miller K8GU wrote:
> Given what
> we know about reciprocal networks, ask yourself, "Why wouldn't
> diffraction be reciprocal?"
>
>
>