Statistical prefiltering for MIMO systems with linear receivers in the presence of transmit correlation

Fading correlation at the transmit antenna array of MIMO systems causes significant performance degradation, especially in case of non-adaptive transmitters and linear receivers. We propose statistical transmit matrix prefiltering schemes for zero forcing and minimum mean squared error receivers that are controlled by knowledge of the transmit correlation matrix only, thus avoiding the need for instantaneous channel state information at the transmitter. To this end, closed-form approximations are derived for average error probability and average mean squared error, respectively, under a flat Rayleigh fading channel assumption. Based on these results, the optimum linear matrix prefilters are derived for a well-known MIMO channel model. Monte-Carlo simulations demonstrate the effectiveness of the proposed statistical prefilters.