EFFECT OF THE SURFACE CORRELATION LENGTH ON MICROWAVE FORWARD REFLECTION FROM A RANDOMLY ROUGH LOSSY DIELECTRIC SURFACE

Abstract

The effect of the surface correlation length on forward reflection from a randomly rough lossy dielectric surface is examined in this paper. At first, the accuracies of the forward reflection models were examined using the moment method (MM). The physical optics (PO) approximation has commonly used to compute the forward reflection coefficient of a randomly rough surface, in which an exponential function of a surface roughness parameter and incidence angle is multiplied to Fresnel reflection coefficient. Another model was proposed by Miller and Brown (MB Model), in which an additional term of modified Bessel function is multiplied to the PO model. The accuracies of those two models were examined based on the moment method solution for various value of kh(rms) where k is wavenumber and h(rms) is the surface RMS(root-mean-square) height. We found that MB model is more accurate for large kh(rms) values. Those models do not include the effect of surface correlation length, where the MM solution shows the effect clearly. Therefore, we examined the effect of correlation length on the reflection coefficients of randomly rough lossy dielectric surfaces using the MM technique.