A new modeling approach for wireless channels with predictable path geometries

We consider a wireless data service characterized by short distances, no shadowing, low power and low antenna heights, deployed in places where a high frequency of potential users is expected, e.g., toll booths, parking lots, intersections, etc. An example of such a service is the infostation concept [1, 2]. Within such a system, we expect to see a well-defined geometry of base-to-user radio paths, as well as a predictable user trajectory; neither of which can be assumed for the wide-area cellular case. This offers the promise of a strong deterministic component of the channel response, in addition to a weaker stochastic component; this is in contrast to a purely stochastic model, as in existing cellular contexts [3, 4]. Here, we present results from measurements we conducted, and the corresponding radio channel models, for three typical outdoor scenarios. Both temporal and spatial channel features are presented. Comparisons between predicted and measured behavior show excellent agreement.