Abstract
This paper is concerned with an algorithm for generating inhomogeneous random rough surface (RRS) that can simulate physical environments such as desert, vegetation ground, sea surface and others, better than homogeneous RRS. First, we discuss statistics of RRS and the direct DFT method which can generate RRS with an arbitrary types of spectrum. Second, we review the essence of the convolution method which is effective and flexible for extension of the algorithm to generate various kinds of RRS. Finally, we propose an algorithm for generating inhomogeneous RRS by modifying the convolution method. Numerical examples are also given for RRS with different types of spectra, correlation length and averaged surface height.