The accuracy of geometric channel modeling methods

This paper addresses the subject which is too often overlooked in the analysis of a geometric model for wireless channel - the axioms corresponding to a priori properties of the channel that is modeled using geometric approach must represent realistic wave propagation. When constructing a model of propagation channel, people often create a substitute by giving an arbitrary shape or form to scatterer distributions based on its intuitive appeal for a certain radio environment. Such models do not necessarily represent actual propagation process and yields deceptive results. The principal goal of this paper is to provide an insight into the underlying connection between any geometric models and the physical propagation process they signify. This is achieved through two parameters: the contour mean probability density function and the effective path length. With these two metrics, two heuristic rules are proposed to provide the missing link between the canonical models and the physical channel. We further revisit a few widely used geometric models in macro- and micro-cellular environments based on the unified analysis.