Power control for space-time-coded mimo systems with imperfect feedback over joint transmit-receive-correlated channel

In this paper, power adaptation strategies for minimizing bit error rate (BER) subject to an average power constraint for orthogonal space-time-coded multiple-input-multiple-output (MIMO) systems with mean and covariance feedback over joint transmit-receive-correlated channels are presented. We first develop an optimal spatial-only power control (PC) based on an approximated BER bound, which can include the mean feedback only and covariance feedback only as special cases. Using a newly defined fuzzy signal-to-noise ratio (FSNR) criterion, a suboptimal spatial PC scheme with a closed-form expression is derived. This FSNR-based power adaptation not only has less computational complexity but has a BER performance close to that of the optimal spatial-only PC scheme as well. Second, the degrees of freedom in temporal and spatial domains are exploited to derive a joint spatiotemporal PC (S-T-PC) with the use of the FSNR strategy to minimize the average BER subject to a long-term average power constraint with mean and covariance feedback. Simulation results show that the proposed two spatial-only PC schemes can provide a BER lower than that of equal power allocation and have almost the same performance as the existing spatial-only optimal scheme but with lower computational complexity. Moreover, they outperform the PC schemes with mean or covariance feedback only. Most importantly, the new joint S-T-PC scheme substantially outperforms all the spatial-only PC schemes.