Optimization for UAV-mounted active RIS assisted multi-user communication in the presence of multiple eavesdroppers^☆

This paper explores a UAV-mounted active Reconfigurable Intelligent Surface (aRIS) network designed to enhance secure downlink communication for multiple users while mitigating the impact of multiple Eavesdroppers (EVs). The focus is on optimizing the UAV’s trajectory, the Base Station’s (BS) transmit beamforming, and the power-Amplified Programmable Reflecting Elements (APREs) of the aRIS to maximize the minimum secrecy rate in the presence of EVs. This is a complex non-convex problem due to multiple optimization variables, high-dimensional matrix operations, and log-determinant objective functions, which makes it challenging to solve. Hence, a Successive Convex Approximation (SCA)-based optimization strategy is developed to efficiently solve the subproblems related to the UAV’s trajectory, aRIS’s APREs, and BS’s beamforming. By leveraging slack variables and approximation techniques, we solve the nonconvex subproblems by a sequence of convex subproblems. Simulation results demonstrate that the proposed UAV-aRIS network significantly outperforms its passive RIS counterpart in improving communication security, highlighting the effectiveness of the optimization strategy.