Throughput Analysis of Adaptive BIC-MIMO-OFDM Systems

The noncoherent capacity of multi-input multi-output orthogonal frequency division multiplexing (MIMO-OFDM) systems is first derived. Then, an adaptive bit-interleaved coded modulation (BICM) scheme for the MIMO-OFDM system that exploits transmitter beamforming and variable-rate-variablepower (VRVP) QAM is proposed to enhance the throughput in a time-varying environment. The throughput of the variable-rate (VR) scheme is proved to have the same pre-log factor as the lower bound of noncoherent capacity. Furthermore, it is proved that the VR scheme has the same diversitymultiplexing tradeoff behavior as MIMO systems with random Gaussian codewords. The effect of imperfect channel station information (CSI) on the effective channel signal-to-noise ratio (SNR) is explained. It is shown by computer simulation that adaptive BICM outperforms adaptive trellis-codedmodulation (TCM) when outdated channel estimation is used. The throughput of multiuser downlink is analyzed and OFDMA with opportunistic scheduling outperforms OFDM-TDMA scheduling with a constant enhancement that is dependent on the user number and the antenna number but independent of SNR.