Large system analysis of power minimization in multiuser MISO downlink with transmit-side channel correlation

This paper investigates a joint power control and beamformer design in a correlated multiuser MISO downlink to minimize the weighted transmit power. We firstly review a distributed fixed-point iterative algorithm. This algorithm, though converging to the optimal solution, requires instantaneous power iteration. In order to reduce the implementation complexity, we leverage tools from random matrix theory to derive an asymptotically optimal solution requiring only statistical channel information. In addition, the effect of transmit-side channel correlation on the large system analysis is studied, and a distributed algorithm computing the asymptotically optimal power is proposed. In this algorithm, the asymptotically optimal power is updated according to channel statistics, and thus no instantaneous power update is needed for computing the asymptotically optimal beamformer.