Abstract: Unmanned aerial vehicles (UAVs) have found important applications in wireless communications and internet of things (IoT) due to their advantages of on-demand and swift deployment, high mobility, and high-quality line-of-sight communication links with ground users. In this paper, a UAV-enabled data collection system is studied, where a UAV collects data from multiple sensors and relays the collected data back to a fusion center. In order to maximize the end-to-end throughput of the data collection system, an optimization problem is formulated to jointly optimize the transmit power and bandwidth of the sensors, as well as the transmit power and flight trajectory of the UAV.The optimization problem is subject to the minimum data transmission constraint, the information-causality constraint, the average and peak transmit power constraint, the bandwidth allocation constraint, and the UAV mobility constraint, and this is a non-convex optimization problem which is difficult to solve. To tackle such a difficulty, this paper proposes an efficient alternative optimization algorithm based on the block coordinate descent method and the successive convex optimization method. The algorithm divides the original problem into a transmit power and bandwidth optimization sub-problem; and a UAV trajectory optimization sub-problem.These two sub-problems are both transformed into convex optimization problems by introducing slack variables and applying the first-order Taylor expansion method,then they can be solved iteratively and alternately. Computer simulation results show that the proposed algorithm can strike a balance between the data collection links and the data forwarding links, and can significantly improve the end-to-end throughput of the system, as compared to other three benchmark schemes, which demonstrates the necessity of optimizing power, bandwidth and flight trajectory jointly.