Passive Precoding and Power Allocation for Energy Efficient Reconfigurable Intelligent Surface Enhanced Millimeter Wave MU-MISO

We consider narrowband downlink millimeter wave (mmWave) multiple-user multiple-input single-output (MU-MISO) communication assisted by several reconfigurable intelligent surfaces (RISs). Hybrid precoding is performed instead of conventional fully digital precoding to reduce the number of expensive and power-hungry radio frequency (RF) chains. Passive precoding. achieved by designing the reflecting coefficients of RIS, is performed together with transmit precoding and power allocation among different users to maximize energy efficiency of the system. By considering the use of zero-forcing (ZF) fully digital transmit precoder, the energy efficiency maximization problem is reduced to solving only passive precoding and power allocation which is achieved through alternating optimization technique. The passive precoding is attained through greedy optimization in phase, whereas power allocation is obtained by the method of Lagrange multipliers using Karush-Kuhn Tucker (KKT) conditions. The energy efficient passive precoding is further advanced through switching off of individual RIS elements accomplished through a low-complexity algorithm. Finally, hybrid ZF precoder is designed from the fully digital precoder. We show, through simulation results, that the proposed design delivers performance better than the existing solution, while entailing lower complexity.