Optimum opportunistic beamforming based on multiple weighting vectors

In order to improve the throughput of the opportunistic beamforming, we generalize the opportunistic beamforming by using multiple random weighting vectors at each time slot. The base station chooses the best weighting vector and performs the opportunistic beamforming with this optimum vector. For the case of equally strong independent fast Rayleigh fading channels, we analytically approximate the throughput of the proposed scheme and obtain the optimum number of random weighting vectors per time slot. Numerical results demonstrate that the proposed scheme considerably improves the throughput compared to the conventional opportunistic beamforming for a low/realistic number of users.