Cluster Identification and Properties of Outdoor Wideband MIMO Channel

In this paper, the statistical analysis of cluster is presented based on the outdoor wideband multiple-input multiple-output (MIMO) channel measurements to facilitate IMT-Advanced system design. Clusters are found in multi-dimensional space, i.e., on the azimuth of arrival-azimuth of departure-elevation of departure-delay domain. We identify clusters with an automatic cluster identification algorithm and a new method is developed to improve the algorithm in terms of convergence rate and accuracy, which can also be used in calculating the angular spread (AS) to avoid the ambiguity caused by the origin of the coordinate system. All cluster parameters are described by a set of probability density functions (pdfs) derived from the measured data. The cluster numbers are well fitted with Poisson distribution plus a minimum number of clusters, and both the delay spread and angular spread exhibit the Lognormal distribution. It is concluded that the elevation angle should not be neglected when terminals are rounded by rich scatterers in outdoor scenario. We also find that clusters with smaller delays have in general a higher cross-polarization discrimination (XPD) than that with larger delays.