Adaptive downlink beamforming for multiuser MISO systems combined with user selection

We consider adaptive downlink beamforming strategy for multiuser Multi-Input Single-Output(MISO) systems. It is well known that the non-linear dirty paper coding (DPC) can entirely eliminate the interference of the other users, but it is computationally prohibitive. The Zero-Forcing Beamforming(ZFBF) is an filternative solution also with high complexity due to the inversion of composite channel matrix. In our proposed adaptive downlink beamforming method, we first estimate the singular value and singular matrices of selected channel matrix, which is obtained using Semi-orthogonal User Selection(SUS). Then the optimal downlink beamforming matrix can be derived using such matrices. Simulation results verify the effectiveness of the proposed scheme.