Multi-UAV Data Collection Framework for Wireless Sensor Networks

TL;DR

This paper presents a comprehensive framework for optimizing multi-UAV data collection in wireless sensor networks, focusing on cost, efficiency, and trajectory planning to improve data gathering performance.

Contribution

It introduces a novel framework that jointly optimizes UAV trajectories and sensor cluster head placement, with near-optimal genetic algorithm solutions and practical guidelines.

Findings

Genetic algorithm achieves near-optimal solutions with 3.5% degradation.

Different trajectory approaches impact data collection efficiency.

Environmental factors and UAV altitude significantly affect performance.

Abstract

In this paper, we propose a framework design for wireless sensor networks based on multiple unmanned aerial vehicles (UAVs). Specifically, we aim to minimize deployment and operational costs, with respect to budget and power constraints. To this end, we first optimize the number and locations of cluster heads (CHs) guaranteeing data collection from all sensors. Then, to minimize the data collection flight time, we optimize the number and trajectories of UAVs. Accordingly, we distinguish two trajectory approaches: 1) where a UAV hovers exactly above the visited CH; and 2) where a UAV hovers within a range of the CH. The results of this include guidelines for data collection design. The characteristics of sensor nodes' K-means clustering are then discussed. Next, we illustrate the performance of optimal and heuristic solutions for trajectory planning. The genetic algorithm is shown to be near-optimal with only $3.5\%$ degradation. The impacts of the trajectory approach, environment, and UAVs' altitude are investigated. Finally, fairness of UAVs trajectories is discussed.