Efficient feedback quantization for orthogonal random beamforming schemes

In multiantenna downlink systems the optimization of the spatial precoding is severely hampered by the amount of feedback. An efficient solution consists in the opportunistic beamforming scheme (OB-MUX) that generates random orthogonal beams and schedules the user with the largest signal-to-interference-and- noise ratio (SINR) over each beam. In this paper we explore the effect of feedback quantization on the performance of OB-MUX scheme. We analytically assess the sum-rate of OB-MUX in terms of quantization strategy, number of users and number of antennas. Since the optimum quantization requires a numerical optimization, we design a practical quantization law that approaches the performance of the optimum scheme at reduced complexity. The proposed strategy permits to extract most of the multiuser diversity gain by using a few (in the limit of one) feedback bits.