A robust AMSE-minimization game in multicell multiuser wireless system

This paper considers a game theoretical approach in designing the multiuser downlink beamformers in practical multi-cell mutiple-input-single-output (MISO) systems, where a channel state information (CSI) model consisting both CSI error and delay is adopted. Under this consideration, each base station (BS) takes the average mean square error (AMSE), which indicates the average difference between the users' expected signal and the real received one, as its optimization object. By regarding the inter-cell interference as colored additional noise, a BS can locally derive a closed-form beamforming scheme by minimizing its AMSE. In this way, the whole system performs a non-cooperative robust game where each BS determines its optimal beamformer in a totally distributed manner. Then the condition guaranteeing the existence and uniqueness of a Nash Equilibrium (NE) in the game is theoretically examined and numerically verified in this paper. Numerical results also verify the advantage in robustness and performance over the non-robust game based on the conventional MSE-based beamformer.