Interference Management in Cellular Full-Duplex MIMO Systems with Statistical CSI

Michael Newinger, Wolfgang Utschick

Abstract: The weighted sum-rate problem for a cellular MIMO network with a full-duplex base station and half-duplex user devices is considered. The system is subject to interference from uplink to downlink users, rendering the optimization problem non-convex. We use interference shaping, which defines sets of admissible interference covariance matrices to separate uplink and downlink into independent sub-problems. While the uplink covariances are optimized to comply with these constraints, a worst case interference optimization has to be performed in the downlink. To limit the channel estimation overhead, it is assumed that only statistical channel state information is available for the interference channels and we propose a scheme with which the uplink can still comply with the shaping sets. For the downlink, we show that the worst case interference problem can be recast into a single convex optimization problem.