On the Beamforming for Multi-Reconfigurable Intelligent Reflecting Surfaces Aided MISO Systems

This paper deals with the problem of the joint design of the phase shift matrices for multi-reconfigurable intelligent reflecting surfaces (RISs), assisting a multi-user multi-input single-output (MU-MISO) system. The RISs are deployed to serve users in the dead zone of the massive MU-MISO system blocked by some obstacle. We propose two passive beamforming techniques, one has a higher complexity since it requires matrix inversion each coherence period, and the other makes use of the channel statistics that is known to change over larger time spans, but its performance decrease as the channel becomes less correlated. Using zero-forcing (ZF) beamforming technique, we show that the proposed methods achieve significant gains compared to random beamforming and a minor gap compared to the more complex exhaustive search method, yet with much less complexity.