Energy-efficient coverage and connectivity maintenance in UAV swarm networks via multi-agent coordination

Abstract

Abstract Time-constrained disaster management with micro-unmanned aerial vehicle (UAV) swarms requires autonomous coordination to ensure effective spatial coverage while preserving reliable ad-hoc network connectivity. The inherent trade-off between coverage and connectivity is compounded by the strict aerodynamic power limitations of micro-UAVs, whose batteries deplete rapidly under erratic maneuvers. Conventional centralized control architectures and standard artificial potential field (APF) algorithms often fail to address this multi-objective problem, leading to fragmented networks or wasted energy due to oscillatory trajectories (jitter). To overcome these limitations, this paper proposes the Energy-Damped Virtual Force (EDVF) algorithm, a fully decentralized, neighbor-aware control law. EDVF employs an adaptive velocity-dependent damping term (analogous to viscous friction) that penalizes unnecessary micro-corrections as drones settle into their topological positions. We formulate a multi-objective optimization model that integrates algebraic connectivity with a non-linear aerodynamic power model based on Blade-Element Momentum Theory. Python-based simulations demonstrate that EDVF outperforms both traditional artificial potential field methods and centralized benchmarks: it maintains approximately 39% percent area coverage – 99% percent of the theoretical maximum for the swarm configuration – guarantees a connected network throughout the mission λ2 > 0, and demonstrably extends swarm operational lifetime by eliminating kinematic oscillations and driving agent velocities asymptotically to zero, thereby removing the cubic aerodynamic drag penalty P_par ∝ ||v||³ that undamped baselines continuously incur. EDVF thus provides a theoretically grounded and scalable solution for maintaining persistent aerial surveillance in stochastic environments.