Statistical beamforming for FDD massive MIMO downlink systems

In this paper, we first present a statistical beam-forming (SBF) design for FDD massive MIMO systems, which is realized only based on statistical channel state information (CSI) and thus can significantly reduce the pilot burden. By fully exploiting the structural feature of statistical CSI in massive MIMO systems, we then derive the closed-form expression of sum-rate for the proposed SBF, revealing the effect of the number of users (receivers) on the sum-rate performance. Furthermore, we utilize the newly derived results to develop an efficient user scheduling algorithm which only requires statistical CSI. Numerical results show that our proposed scheme outperforms the signal-to-leakage-and-noise ratio based user scheduling algorithm. In particular, it is shown that when the number of users is large enough, the proposed SBF transmission scheme is comparable to instantaneous CSI based zero-forcing beamforming scheme for highly correlated channel with reasonable channel estimation error.