Clusterization of measured direction-of-arrival data in an urban macrocellular environment

Abstract

Measured wideband direction-of-arrival data from urban macrocellular environment were clusterized. The clusterization was done by manual inspection. To understand the physical propagation channel, the source points of the extracted clusters were located to the local environment surrounding the mobile station. Some key parameters were calculated for each of the extracted clusters and for the residual power. By combining these parameters with the estimates of the cluster source points, assumptions on the propagation mechanisms of the clusters were drawn. According to the results the most significant sources are building structures above the average building height and corners of the buildings. This result is inline with previous studies in the urban propagation. Only a few significant clusters were found, in this study three clusters out of 14 extracted clusters contributed nearly 60% of the received power, while 19% of the power remained outside the clustering. Most of the clusters follow the Rician distribution with relatively high K-factor, while the residual power was Rayleigh distributed. There was only slight depolarization from the transmitted vertical polarization. One of the clusters was a clear exception with a negative XPR.