Optimal power allocation with partial channel knowledge for MIMO multicarrier systems

Closed-loop schemes optimizing power allocation can be applied if channel status information (CSI) is available at both transmitter and receiver. However, CSI is never ideal and algorithms performance degrades rapidly when CSI is poor. In this contribution a MMSE cost function that takes into account channel uncertainty is proposed for MIMO systems with OFDM modulation. The derived algorithm is robust to channel estimation errors. A closed form solution is obtained for joint transmitter and receiver design that optimally allocates the power over all subcarriers and antennas. Numerical results show the new criterion outperforms other solutions that assume ideal CSI knowledge.