Optimal Real Time Drone Path Planning for Harvesting Information from Moving Sensors in a Wireless Sensor Network

Abstract

We consider a remote sensing system in which a drone collects information from moving wireless sensor cluster heads under a fixed path-length constraint. The objective is to minimize the transmission energy required for communication between the moving cluster heads and the drone, where transmission cost follows a distance-based power law. We formulate the problem as a constrained optimization problem and derive a vector-field differential equation that continuously deforms feasible drone trajectories toward locally optimal solutions while preserving the path-length constraint. To address the dependence of the solution on cluster visitation order, we develop a hybrid optimization framework combining continuous trajectory optimization with multiple iterative reordering strategies based on predicted cluster positions and traveling-salesman heuristics. The proposed methods were evaluated on six classes of Monte Carlo simulation scenarios involving different numbers of cluster heads, target speeds, and allowable drone path lengths. The best-performing algorithms, particularly the FinalXY and FinalUV methods, consistently achieved dramatic reductions in transmission energy compared to static TSP-based routing. For scenarios with 20 moving cluster heads, optimized solutions typically reduced transmission energy to less than 10% of the initial static-ordering solution, while scenarios with 40 cluster heads achieved reductions below 5%. The algorithms also computed near-optimal paths in seconds on standard desktop hardware, demonstrating suitability for real-time adaptive drone path replanning in dynamic environments.

Keywords: Wireless Sensor Network, Drone, Path Planning, Information Collection, Power-Efficient, Extended Lifetime, Optimization, TSP, Gradient Descent