My wild guess (I'm in compter graphics) is that probably not very accurately, since the wavelength of radio waves is so long compared to the environment where they propagate, that wave effects (like diffraction) will play a major role. RT typically sucks in such conditions, since it assumes geometric optics.

thank you for your answer but I want to know exist there any free software or programs in C, C ++, or Java that can help me to model a channel?

It all depends to what level of accuracy you want the results. As Oskar said, ray tracing deals with visible light, or actually straight lines. One empirical model you can program yourself, because it's quite straightforward, would be the extended Hata model:

http://w3.antd.nist.gov/wctg/manet/calcmodels_r1.pdf

This takes into account antenna heights above ground, and building clutter too.

And then, of course, you can use free space propagation, which doesn't take into account anything about obstacles.

http://www.radio-electronics.com/info/propagation/path-loss/free-space-formula-equation.php

That one is even easier to program on your own.

The most comprehensive RF propagation model I know is the Longley-Rice model, but it requires a lot of detail of the surrounding geography. And as often happens with very elaborate models, at least in my experience, the results are not always more reliable than those of simpler models! The Hata model is a good start.

thank you for your answer, you are absolutely right, but I want to model the channel by a deterministic model so it must apply the ray tracing method and add the model in Uniform Theory of Diffraction (UTD).

Can you help me?