Abstract

Air-to-ground interference poses a critical challenge in integrating unmanned aerial vehicles (UAVs) into cellular networks. In downlink scenarios, UAVs can withstand significant interference from co-channel base stations (BSs) due to the guaranteed line-of-sight (LoS) connection with ground users. Our research focuses on power optimization in BSs and applying green energy principles to pave the way for more sustainable and energy-efficient BSs within UAV-integrated wireless systems. To this end, this paper investigates a downlink UAV communication scenario in which the intelligent reflecting surface (IRS) is mounted on the UAV to practically nullify the interference originating from the co-channel BSs. We formulate the IRS beamforming matrix to reduce transmit power by optimizing passive beamforming for IRS elements, incorporating adjustments to phase shifts and amplitude coefficients while considering the positioning of the UAV. The proposed optimization problem is non-convex, and thus a successive convex approximation (SCA) method is adopted to convert all constraints to a quadratic approximation. Simulation results demonstrate that the proposed SCA algorithm provides an efficient transmit power minimization approach with low computational complexity for large IRS since it achieves close-to-optimal performance, and significantly outperforms conventional systems without IRS. In interference scenarios and with different numbers of IRS meta-atoms, the proposed algorithm achieves a power reduction of approximately 8 and 13 dBm, while maintaining the same required signal-to-interference-plus-noise ratio.

Key words: Intelligent reflecting surfaces (IRS), successive convex approximation (SCA), cone programming, unmanned aerial vehicles (UAV).