부산대학교 도서관

This paper exploits the mobility of an unmanned aerial vehicle (UAV) to improve the physical layer (PLS) security against potential eavesdropping scenarios, where the sensitive data transmitted from a ground base station (GBS) to multiple information receivers (IRs) is easily eavesdropped by multiple non-colluding eavesdroppers (Eves). Specifically, the UAV is employed as a friendly jammer to transmit artificial noise to degrade the wiretap channels. By leveraging the UAV's mobility and the possibility of establishing line-of-sight links, the UAV can move close to the Eves and opportunistically jam them. Especially, two practical factors, namely, the imperfect location information of Eves and the existence of no-fly zones are considered. Since only the partial position information of the Eves can be known, as the worst case, the minimum average secrecy rate of IRs is robustly optimized by jointly designing the trajectory and jamming power of the UAV, and the transmission power of the GBS to ensure the security of all IRs. However, the formulated problem with coupled variables is non-convex and intractable. As such, an efficient alternative iterative suboptimal algorithm is proposed to solve this problem by applying the block coordinate descent, successive convex approximation, slack variables, and $\mathcal {S}$-Procedure. Numerical results reveal that the proposed robust algorithm can substantially improve the max-min average secrecy rate, and the effectiveness of the proposed algorithm is further verified by comparing it with three benchmark schemes.