An Empirical Investigation of Multi-Path Clusters in an Outdoor MIMO Propagation Environment

This paper investigates the cluster dynamic behaviors in an outdoor propagation environment to facilitate the outdoor channel modeling and simulation. To identify the relevant clusters from measured channel data in joint spatial-temporal domain, an automatic cluster identification algorithm has been employed. By introducing a closed loop scheme and a multi-dimension filter, significant improvements on the algorithm performance can be obtained. Furthermore, based on analysis on the cluster identification results, this paper proposes two stochastic models for cluster time-variant characteristics including cluster lifespan and cluster spatial-temporal evolution. Empirical examples reveal the applicability and accuracy of the models of which can be made use in related simulations to imitate the continuous evolution of realistic channels.