Path planning for UAV harvesting information from dynamical wireless sensor nodes at sea

Abstract

A system of several wireless sensor nodes and one unmanned aerial vehicle (UAV) is considered in this research. The nodes are only floating and drifting with the sea stream. The UAV will be operating as a data mule to gather sensing information from wireless sensor nodes. Unlike prior studies, this paper addressed a realistic ocean model for the nodes movements which will be the references to the Kalman Filter (KF) in estimating for the nodes’ positions. Simulation results are evaluated for an optimal flight-able path for the UAV under several constraints by particle swarm optimization (PSO). Specifically, the deviation between the estimated positions and the referenced positions, total energy consumption by the sensors network, data rates between UAV and the nodes, flight time for the UAV, and frequency of visiting the nodes by the UAV will be considered for optimization. The systems performances will be evaluated based on these scenarios: a) an ideal and unrealistic scenario where the UAV follows the nodes continuously; b) a realistic case where the UAV only flies periodically. Discussions and solutions were also addressed for the situations when the deployed nodes are more significantly separated than the cases simulated in the paper.