Active and passive beamforming based on RWMMSE and Gradient Projection for IRS-aided MU-MISO Systems

The intelligent reflecting surface (IRS), a planar array consisting of many low-cost configurable reflecting elements, offers many new possibilities, such as creating a virtual line of sight, interference cancellation, wireless power transfer, etc. To efficiently achieve these goals, the real-time reconfiguration of its elements, which is still challenging, needs to be as simple as possible. Prior works for downlink IRS-aided multi-user multiple-input single-output (MISO) systems, which are noted to have high complexity, are extended in this work by incorporating them with novel techniques that lower their complexity. In doing so, we create a weighted sum rate problem and adopt a reduced weighted minimum mean square error (RWMMSE) algorithm, which is an advanced low-complexity version of the weighted minimum mean square error (WMMSE) scheme to optimize the active precoders, whereas a gradient projection (GP) algorithm, whose procedure of choosing the step size is redesigned to avoid the high-complexity iterative processes, is adopted for optimization of passive beamformers. Finally, the presented numerical results demonstrate the effectiveness of our proposed scheme in attaining comparable performance to the benchmark schemes with a significant complexity reduction.