antenna-theory.com

[jacobkusk]

Hey
Thanks for an excellent site! I'm totally newbie in the field of electromagnetic waves and antenna theory. Have a question to this site.

Imaging a isotropic antenna dropped into a medium of infinite size. This medium could be water, corn, plastic balls, flour etc. The perfect antenna is transmitting with a constant energy and a constant frequency.

How can I estimate the propagation loss in this case?

As i can understand from theory is that the signal will be absorbed, scattered, (reflected) and change wave length.

Is there a database or something with absorption and scatter coefficients for different mediums?

Sorry, if my English is not perfect.

[Schubert]

First, you have to assume the medium is homogenous. This means the material surrounding the antenna is relatively uniform given distances "relative to a wavelength" at the frequency of transmission. This means, for instance, that a material such as plastic balls would be homogenous if the ball radius is much less than a wavelength, and not homogeneous if the balls were roughly the same size or larger than a wavelength.

Now, if you have a homogenous medium, what you need to know is the dielectric constant (epsilon or permittivity) of the material. If it is magnetic, you need the permeability (mu) of the material. Finally, if the material has any electrical conductivity (sigma), you'll need to know that as well.

Then you can go to the Friis Transmission formula:
http://www.antenna-theory.com/basics/friis.php

Assuming that the material is non-conductive, the path loss can be modified by multiplying the standard equation by 1/sqrt(eps*mu). This is because the speed of light is slower in a non-vacuum material, with more energy in the associated fields, leading to a greater path loss.

If the conductivity is non-zero, you'll also need an exponential decay term in there to account for the resistive losses in the medium that will further increase the path loss.

[jacobkusk]

Schubert: Thank you for the answer. The 1/sqrt(eps*mu) is that based on your experience or do you have a reference where I can read more about it?

[Schubert]

Ok, now I'm not sure. In the case of a non-zero conductivity, there will certainly be increased energy lossed in the resistive material.

But now I'm thinking, if the power transmitted is Pt, then in a lossless medium the power density at a distance R must be Pt/(4*pi*R^2)

Hence, I'm not sure the pathloss (which relates power transfer) as a function of R should be a function of epsilon and mu.

Sorry, I have to think about it more.