Collaborative relay-based multiuser beamforming in cellular systems

The performance of downlink transmissions can be improved by using relays that can form a distributed virtual multiple-input multiple-output (MIMO) beamforming system. In this paper, we study a collaborative downlink transmission scheme for a cellular system which consists of a base station (BS), multiple amplify-and-forward (AF) relay nodes, and multiple users. We propose a new algorithm to determine the signal forwarding weights for both transmit power and signal phase with an objective of minimizing power consumption of the relay nodes while guaranteeing the receive signal-to-noise ratio (SNR) at multiple users for the collaborative relay beamforming. We prove that the beamforming solution obtained by a semidefinite relaxation (SDR) technique is optimal with the proposed algorithm. Numerical results validate our theoretical analysis and demonstrate that the proposed optimal scheme can effectively reduce the power consumption of relay nodes compared with other beamforming approaches.