Variable-rate adaptive modulation in MIMO systems exploiting multiuser diversity

An adaptive modulation technique for multiuser MIMO systems with multiuser diversity is proposed. This technique maximizes the sum of the instantaneous bit rate under a target BER constraint by feeding back sub-channel SNRs only. By generating a random transmit beamforming matrix with a water-filling configuration, each user estimates and feeds back the effective SNRs for its sub-channels. With each user´s sub-channel SNRs, the transmitter then schedules the transmission to a particular one selected by a proportional fair scheduling algorithm. Finally, adaptive modulation is applied to each sub-channel of the selected user based on the BER constraint. Simulation results show that the throughput of the proposed method converges to that of eigen-beamforming when the number of users in a cell is large.