Statistical distribution of path interarrivaltimes in indoor environment

Abstract

Interarrivaltimes of paths are investigated with respect to their statistical distribution on the basis of a total amount of 5606 impulse responses measured in 63 different scenarios in an indoor environment. For the detection of paths, an algorithm denoted as "path tracking" is applied, which enables the determination of physically reasonable paths. Evaluation is done by the examination of the frequency distributions of errors when fitting theoretical distribution functions to measured data. An exponential interarrivaltime distribution, as occurring for a standard Poisson process, shows poor correspondence, whereas a Weibull distribution turns out to be very well applicable. A plausible explanation for the applicability of the Weibull distribution for describing the distribution of interarrivaltimes, which was missing up to now, is also given.