Novel Power Adaptation Strategy for Space–Time-Coded Multiantenna Systems With Imperfect Channel State Information

This paper presents a strategy for allocating power in both space and time for beamforming space-time-coded multiantenna systems with imperfect channel state information at the transmitter (CSIT) over flat and quasi-static Rayleigh fading channels. We first develop a closed-form spatial-only (S-O) power allocation scheme based on a newly defined compressed signal-to-noise ratio (CSNR) criterion. The derived power allocation formula has a parameter (compression factor) that can be used to minimize the conditional bit error rate (BER). The BER performance of the new CSNR-based power allocation scheme is almost identical to the previous optimal S-O power allocation schemes, whereas its computation is more efficient. Using the CSNR-based power adaptation, we derive a joint spatio-temporal (S-T) power allocation method to minimize the average BER subject to a long-term (time) average of the transmit power constraint, which provides a large BER performance gain over the S-O power strategies and the existing S-T method. The practical issue of peak power constraint is also addressed.