In this paper, we consider a scenario where an unmanned aerial vehicle (UAV)collects data from a set of sensors on a straight line. The UAV can eithercruise or hover while communicating with the sensors. The objective is tominimize the UAV's total aviation time from a starting point to a destinationwhile allowing each sensor to successfully upload a certain amount of datausing a given amount of energy. The whole trajectory is divided intonon-overlapping data collection intervals, in each of which one sensor isserved by the UAV. The data collection intervals, the UAV's navigation speedand the sensors' transmit powers are jointly optimized. The formulated aviationtime minimization problem is difficult to solve. We first show that when onlyone sensor node is present, the sensor's transmit power follows a water-fillingpolicy and the UAV aviation speed can be found efficiently by bisection search.Then we show that for the general case with multiple sensors, the aviation timeminimization problem can be equivalently reformulated as a dynamic programming(DP) problem. The subproblem involved in each stage of the DP reduces to handlethe case with only one sensor node. Numerical results present insightfulbehaviors of the UAV and the sensors. Specifically, it is observed that theUAV's optimal speed is proportional to the given energy and the inter-sensordistance, but inversely proportional to the data upload requirement.
展开▼
