[Zurück]

N. Czink, P. Cera, J. Salo, E. Bonek, J.P. Nuutinen, J. Ylitalo:
"Improving clustering performance using multipath component distance";
Electronics Letters, 42 (2006), 1; S. 33 - 35.

The problem of identifying clusters from MIMO measurement data is addressed. Conventionally, visual inspection has been used for cluster identification, but this approach is impractical for a large amount of measurement data. For automatic clustering, the multipath component distance (MCD) is used to calculate the distance between individual multipath components estimated by a channel parameter estimator, such as SAGE. This distance is implemented in the well-known KMeans clustering algorithm. To demonstrate the effectiveness of the choice made, the performance of the MCD and the Euclidean distance were compared by clustering synthetic data generated by the 3GPP spatial channel model (SCM). Using the MCD significantly improved clustering performance.

http://aleph.ub.tuwien.ac.at/F?base=tuw01&func=find-c&ccl_term=AC06586026

http://publik.tuwien.ac.at/files/pub-et_11004.pdf