Robust Optimization Framework for Ground Coverage in Aerial Sensor Networks

Abstract

Sensors play a critical role in environmental monitoring, but their coverage performance degrades significantly under spatial uncertainty. This article proposes a robust optimization framework for maximizing ground coverage in aerial directional sensor networks subject to sensor displacement. Each aerial sensor projects a truncated sector on the ground, parameterized by its altitude, field of view, and orientation. To explicitly capture robustness against positional uncertainty, we adopt the radius of robust feasibility (RRF) as a quantitative measure of tolerance to worst-case perturbations. The RRF formulation for aerial sensor networks is embedded directly into the coverage maximization problem, ensuring feasibility under bounded uncertainty. The resulting worst-case coverage problem is nonconvex and NP-hard; therefore, a distributed greedy orientation algorithm based on Voronoi partitioning is applied to adjust sensor orientations using only local information, while directing coverage toward high-impact regions. Simulation results demonstrate that the proposed method consistently preserves robust coverage across complex terrain, varying parameters and uncertain operating conditions, highlighting its practical significance for aerial sensing applications.