Hi WhoJuntao: you've got two main effects here:
1/ if the obstacle is << wavelength then the wave is only slightly perturbed by its presence and tends to wrap around it.
2/ If the obstacle is a dielectric with low loss and with embedded mesh or grid of conductors then when the wave length...
Hi - it sounds as though you have a flat fading channel model where a new set of channel gain and phase shift characteristics are calculated only occasionally (i.e. the fade charactristics change only slowly compared with the signal that you want to receive so the description of the fading only...
As pointed out the light pressure does tend to contribute to the orbital perturbations , what is often overlooked is that the microwave power beam has a 'light pressure' all its own!