MIMO linear precoder design with non-ideal transmitters

We investigate the linear precoder design problem for multiple-input multiple-output (MIMO) channels under non-ideal transmitter hardware. We consider two different non-ideal hardware models: i) an additive noise model in which the level of the noise at an antenna is proportional to the signal power at that antenna, ii) an additive precoder error model. We focus on the problem of minimizing mean-square error at the receiver under transmit power constraints at the transmitter. For the first hardware impairment model, this scenario leads to a non-convex formulation for which we propose a block-coordinate descent technique. The proposed method has a convergence guarantee and provides rank-constrained solutions. For the second model, analytical expressions for the optimum designs are provided. We compare the performance of our hardware impairment aware designs with that of designs developed with ideal hardware assumptions. Our results suggest that significant gains can be obtained by the proposed designs for sufficiently high signal-to-noise ratio values.