Subspace Beamforming and Scheduling for the Multiuser MIMO Broadcast Channel

Abstract:
Recently, capacity achieving transmit strategies have been derived for the MultiUser (MU) Multiple-Input Multiple-Output (MIMO) broadcast channel. Contrary to these optimal approaches, the heuristic Successive Encoding and Successive Allocation Method (SESAM) which still performs close to optimum, is computationally more efficient because it avoids the iterative solution of the underlying optimization problem. In this paper, we present a modification of the SESAM whose computational complexity is even further reduced and which requires less Channel State Information (CSI) at the transmitter. Both is achieved by restricting the beamforming vectors for each user to predefined subspaces known at the transmitter and the corresponding receivers. Simulation results of a MU MIMO Orthogonal Frequency Division Multiplexing (OFDM) system will show that the performance loss can be kept moderate although the computational complexity and CSI is reduced tremendously.