This paper considers a dual-unmanned aerial vehicle (UAV) enabled secure communication system under the probabilistic line-of-sight (LoS) channel model. Specifically, one UAV transmits confidential information to multiple ground nodes while the other UAV cooperatively sends jamming signals to confuse a potential ground eavesdropper. We maximize the minimum worst-case average (expected) secrecy rate of the considered system, by jointly optimizing the communication scheduling, robust three-dimensional (3D) trajectories, transmit/jamming powers of both the UAVs for a given flight duration. However, the formulated problem is difficult to obtain a globally optimal solution due to the imperfect location information of the eavesdropper and its non-convex constraints. As such, we first derive an approximated (expected) secrecy rate with high accuracy, based on which we solve it by applying the block coordinate descent and successive convex approximation techniques to obtain its suboptimal solution. Simulation results show that the superiority of the proposed algorithm in urban environments, compared to other benchmarks.

本文考虑在概率视距 (LoS) 信道模型下启用双无人机 (UAV) 的安全通信系统。具体来说，一架无人机向多个地面节点传输机密信息，而另一架无人机则协同发送干扰信号以迷惑潜在的地面窃听者。我们通过联合优化通信调度、稳健的三维 (3D) 轨迹、两个无人机在给定飞行持续时间内的发射/干扰功率，最大化所考虑系统的最小最坏情况平均（预期）保密率。然而，由于窃听者的位置信息不完善及其非凸约束，公式化的问题难以获得全局最优解。因此，我们首先推导出高精度的近似（预期）保密率，在此基础上，我们通过应用块坐标下降和逐次凸逼近技术来求解它，以获得其次优解。仿真结果表明，与其他基准相比，该算法在城市环境中的优越性。