Transmit Optimization for MIMO-OFDM With Delay-Constrained and No-Delay-Constrained Traffic

This paper studies low-complexity power control schemes to approach the capacity limits of multiple-input multiple-output orthogonal-frequency-division-multiplexing (MIMO-OFDM) channels under the assumption that perfect channel state information (CSI) is available at both the transmitter and the receiver. Two notions of channel capacity limit are considered, namely, the ergodic capacity for variable-rate transmission of no-delay-constrained (NDC) traffic and the delay-limited capacity for constant-rate transmission of delay-constrained (DC) traffic. Let M be the minimum between the number of transmit and receive antennas and N the number of subcarriers, we propose a subchannel grouping (SCG) technique, which divides the total number of MN space-frequency MIMO-OFDM subchannels into M grouped channels, each exhibiting a different fading statistics. A class of multitarget power control schemes are then proposed to allocate transmit power and rate to each grouped channel via different statistical water-filling (SWF) solutions. More specifically, multitarget constant-power (MT-CP) is proposed for transmission of NDC traffic and multitarget overall channel inversion (MT-OCI) for transmission of DC traffic. Furthermore, we consider the problem of transmission with mixed DC and NDC traffic and present for this case a low-complexity power control scheme which is based on hybrid MT-CP and MT-OCI. Because of SCG and SWF, the proposed schemes reduce significantly the computational complexity required by optimal power and rate allocation in MIMO-OFDM channels. Numerical examples are provided to evaluate the achievable rates of the proposed power control schemes and compare them with the capacity limits of MIMO-OFDM channels