Robust precoding for multicell MIMO downlink systems

The robust linear transceiver design for the more general model of multicell MIMO downlink system with imperfect channel state information (CSI) is discussed in this paper. Our aim is to minimize the total power of the network while the quality of service (QoS) in terms of mean-square error (MSE) for every user should be strictly guaranteed for every channel realization in the uncertain region. The optimization problem is troublesome due to infinite nonconvex MSE constraints. To handle this, we propose an iterative algorithm that performs optimization alternatively by switching between the precoders and decoders. The two subproblems in the algorithm can be recast as semidefinite programming (SDP) problems which can be efficiently solved. Finally, simulation results show that our proposed algorithm can provide guaranteed QoS for users and outperforms single cell designs especially in serious inter-cell interference (ICI) scenario.