Geometry-Based Clustering Characteristics for Outdoor Measurements at 28-30 GHz

This letter presents an analysis of channel characteristics for an outdoor scenario at the frequency band of 28-30 GHz. A geometry-based clustering algorithm is proposed to group the measured multipath components (MPCs) from the interaction points on the surrounding walls or objects according to ray-tracing three-dimensional (3-D) simulations, giving results that are more physically interpretable than the traditional clustering algorithms. The cluster-level characteristics at 15 positions along a predefined route covering both line-of-sight (LoS) and non-LoS (NLoS) scenarios are investigated. Moreover, power contributions of MPCs with multiple bounces (up to seven bounces in NLoS scenario) from the interaction objects are also investigated.