Abstract: In order to maximize the weighted sum rate of the multi-input single-output non-orthogonal multiple access ( NOMA ) communication system assisted by the reconfigurable intelligent surface ( RIS) , a method combining fractional programming and successive convex approximation is proposed to alternatively optimize active and passive beamforming. First, the weighted sum rate problem is transformed into a fractional problem by fractional programming and the signal and noise power are decoupled. Then, the active and passive beamforming are alternately optimized by the Lagrange multiplier method and successive convex approximation, respectively. Thus, the nonconvex problem of maximizing the weighted sum rate is decomposed into multiple convex optimization problems. Finally, the optimal solution of active beamforming and the suboptimal solution of passive beamforming are obtained. The simulation results show that the proposed method can effectively solve the beamforming problem in the RIS-NOMA communication system and users can obtain an ideal rate increase.

Key words: reconfigurable intelligent surface, non-orthogonal multiple access technology, beamforming,   alternating optimization method, weighted sum rate