Per-Antenna Power Constrained Cooperative Beamforming in Distributed Relay Networks

We consider a bent-pipe, coherent distributed relay network that functions as an antenna array employing transmit beamforming to coherently combine a payload signal and reject interference. A traditional unconstrained minimum mean-square error (MMSE) beamformer can be employed to construct the beamformer, but the unconstrained optimization disregards the power limitations of individual elements. Deploying such a beamformer requires lowering the transmission power to avoid exceeding power limits, which reduces the achievable signal-to-interference-plus-noise ratio (SINR). In this paper, we develop a per-element power-constrained MMSE beamformer that simulta-neously maximizes the SINR. We formulate this construction as a biconvex program and solve it using interior point and alternating proximal method. We demonstrate through simulations that this method produces beamformers that achieve signal-to-noise ratio (SNR) improvements and interference rejection (INR) capabilities comparable to unconstrained MMSE beamformers but with improved achievable SINR.