LAUSR

We investigate the importance of weak clusters when modeling a wireless massive MIMO channel. We do this by studying the influence of densely spaced terminals and the number of base-station antennas for a zero-forcing precoded massive MIMO system. In particular, we focus on the influence on the correlation and imbalance between the signals at the terminal antennas, the effective channel-gain, the eigenvalue distributions and the number of clusters. The study is based on measured radio-channels from terminal prototypes with integrated antennas connected to a massive MIMO testbed. We further evaluate the advantage of using block-diagonalized zero-forcing compared to conventional zero-forcing in a massive MIMO system. Unexpectedly, terminals with low antenna envelope correlation coefficient may benefit significantly from block-diagonal zero-forcing in a massive MIMO system. The main conclusion is that weaker clusters are important when modeling multi-user scenarios.