3D Trajectory Design for UAV-Enabled Data Harvesting in Probabilistic LoS Channel

In this paper, we consider a UAV-enabled wireless sensor network (WSN), where a UAV is dispatched to collect data from multiple sensor nodes (SNs) at known locations within a given duration. In urban areas, the signal propagation between the UAV and SNs can be occasionally blocked (or at least severely attenuated) by high and dense buildings. To address this issue, we first establish a probabilistic line-of-sight (LoS) channel model for a Manhattan-type city by using simulation and data regression methods, which is shown in the form of a generalized logistic function of the UAV-SN elevation angle. Based on the obtained channel model, an off-line optimization problem is then formulated to maximize the minimum expected (average) data-collection rate from all SNs by jointly designing the UAV three-dimensional (3D) trajectory and transmission scheduling of SNs. Since the expected rate is a highly complex function of the 3D UAV trajectory, we approximate it by a tractable lower bound based on the dominant rate in the LoS channel state. The resultant optimization problem is still non-convex and thus difficult to solve. As such, we further propose an efficient algorithm to solve it sub- optimally by applying the techniques of block coordination descent and successive convex approximation. Simulation results are presented to demonstrate the effectiveness of the proposed algorithm and reveal useful properties of the optimized 3D UAV trajectory for data harvesting in WSNs.