Efficient matrix volume control to maximize correlated MIMO channel capacity

Within this contribution we have focused on selec- tion of number of active RF chains under particular channel correlation and SNR dynamics. Study is based on maximizing the volume of the submatrix to be selected. We have proposed a maximum volume submatrix selection (MVC) procedure, that determines the cardinality of number of transmit antennas to be used followed by low complexity greedy subset selection algorithm. Cardinality of the transmit antennas is determined such that mutual information is maximized under particular modulation order. We have observed that for low correlation setups higher number of RF chains are required as compared to high correlation dynamics. Monte carlo simulations validate that performance of our proposed MVC algorithm is very close to optimal water-filling solution for capacity maximization.