Estimation of the Correlation Properties of Large Scale Parameters from Measurement Data

Interdependences of radio-channel model parameters, being observed in some measurement data, should be reproduced by the model. For that purpose the statistical correlation is often used. From the same wideband multiple-input and multiple- output (MIMO) channel data, estimated correlation between large-scale propagation parameters (e.g. delay spread and shadow fading) could be different due to non-uniqueness of post-processing procedure. In this paper, through examining a set of measurement data, we studied the impacts of differently parameterized post-processing procedures of measurement data on the auto- and cross-correlation properties of large scale parameters. We focus on the single parameter in postprocessing procedures: different margins of noise cutting level. The measurement data are gathered in a public hotspot bridge-to-car highway LOS scenario in Ulm, Germany.