Active RIS Aided Multi-User Broadcast System: A Novel Penalty-Based Moth-Flame Optimization Algorithm for the Optimal Location of RIS

In this paper, we consider an active reconfigurable intelligent surface (ARIS) aided multi-user broadcast system, and given fixed electromagnetic coefficients at transmitter and ARIS, we focus on the optimal location deployment of ARIS in a two dimensional feasible area so as to enhance the sum-rate performance of all users. To deal with this complicated nonconvex problem, we firstly approximate the problem by omitting the non-line-of-sight (NLoS) part in the reflection channels of RIS that account for a small proportion of signal power. And then, we propose a novel penalty based moth-flame optimization algorithm to solve the approximated problem by optimizing the coordinates of ARIS that are highly coupled in both the large-scale fading and LoS components. Numerical results have verified that the proposed MFO algorithm is with monotonic convergence, and although the NLoS components is not included in the algorithm, a significantly better performance gain of the sum-rate can be achieved given the newly updated location of ARIS with re-estimated NLoS components compared with the benchmark schemes.