We analyze the end-to-end performance of an unmanned-aerial-vehicle (UAV)-assisted data ferrying network where the UAV serves as a data ferry between the source base station (BS) and multiple destination receivers. We evaluate the end-to-end reliability both with and without packet retransmission technique called automatic repeat request (ARQ), bit error probability (BEP), energy-efficiency, and transmission outage probability. We consider line-of-sight (LoS) and non-line-of-sight (NLoS) transmissions in both the data loading and delivering links and model them with the Rician and Rayleigh fading channels, respectively. We derive tractable approximations for the derived SNR outage results and demonstrate their application in optimizing the distance that a UAV should travel in order to balance the energy-coverage trade-offs. We formulate two different optimization problems and convexify them to solve for the optimal ferrying distance, i.e. (i) outage-constrained energy minimization and (ii) energy-constrained SNR outage minimization. Closed-form optimal solutions are obtained for the second problem. In addition, we formulate a bi-objective optimization problem in order to minimize SNR outage and energy consumption with desired SNR outage probability constraints. The objective function is then reformulated using difference of convex functions (DC) and solved using a DC algorithm. Numerical results validate the derived expressions and show a comparison of the obtained solutions with the solutions obtained from exhaustive search. Insights related to the impact of LoS Rician and NLoS Rayleigh fading channels as well as the optimal ferrying distance are obtained considering a variety of objective functions.

中文翻译：

---

无人机辅助无线数据传输的端到端能源效率和可靠性

我们分析了无人机 (UAV) 辅助数据传输网络的端到端性能，其中 UAV 作为源基站 (BS) 和多个目的地接收器之间的数据传输。我们评估了使用和不使用称为自动重复请求 (ARQ)、误码概率 (BEP)、能源效率和传输中断概率的数据包重传技术的端到端可靠性。我们考虑了数据加载和传输链路中的视距 (LoS) 和非视距 (NLoS) 传输，并分别使用 Rician 和 Rayleigh 衰落信道对它们进行建模。我们为派生的 SNR 中断结果导出了易于处理的近似值，并展示了它们在优化无人机应行进的距离以平衡能量覆盖权衡方面的应用。我们制定了两个不同的优化问题并将它们凸化以解决最佳传输距离，即（i）中断约束能量最小化和（ii）能量约束 SNR 中断最小化。对于第二个问题，得到了封闭形式的最优解。此外，我们制定了一个双目标优化问题，以便在具有所需 SNR 中断概率约束的情况下最小化 SNR 中断和能源消耗。然后使用凸函数差 (DC) 重新制定目标函数，并使用 DC 算法求解。数值结果验证了导出的表达式，并显示了获得的解决方案与从穷举搜索获得的解决方案的比较。