Vehicle-to-Infrastructure (V2I) networks have a wide application prospect for providing vehicles with reliable road safety and infotainment services. This paper studies an Unmanned Aerial Vehicle (UAV)-assisted vehicular network, in which a UAV with cache is dispatched to communicate with moving vehicles and base stations. We study the throughput maximization problem by optimizing the power distribution and trajectory planning of the UAV subject to practical mobility constraints and the information-causality constraint. To solve the non-convex problem, we break it into two sub-problems and propose an iterative algorithm that jointly applying slack variables and sequential optimization. Due to the limited on-board energy of UAVs, improving the energy efficiency (EE) of the UAV is of great significance. To this end, we propose an efficient algorithm by merging fractional programming and sequential optimization. Moreover, we set two special schemes as benchmarks to measure the performance of the proposed algorithms. Numerical results show that the application of cache and the proposed algorithms notably enhance the throughput and EE of the UAV-assisted V2I network.

车辆到基础设施（V2I）网络在为车辆提供可靠的道路安全和信息娱乐服务方面具有广阔的应用前景。本文研究了无人飞行器（UAV）辅助的车载网络，其中分配了具有缓存的无人飞行器以与移动的车辆和基站进行通信。我们通过在实际移动性约束和信息因果关系约束下优化无人机的功率分配和轨迹规划来研究吞吐量最大化问题。为了解决非凸问题，我们将其分为两个子问题，并提出了一种迭代算法，该算法联合应用松弛变量和顺序优化。由于无人机的机载能量有限，因此提高无人机的能效（EE）具有重要意义。为此，通过合并分数规划和顺序优化，我们提出了一种有效的算法。此外，我们设置了两个特殊的方案作为基准来衡量所提出算法的性能。数值结果表明，缓存的应用和所提出的算法显着提高了无人机辅助V2I网络的吞吐量和EE。