RAMNA: A Resource-Aware Algorithm for Maximizing Availability in Flying Ad-Hoc Networks
TL;DR
RAMNA is introduced, a resource-aware optimizing algorithm designed to maximize network availability by autonomously repositioning UAVs at runtime, and contributes to the self-healing and self-optimization properties required for resilient, long-lived flying ad-hoc networks operating under energy uncertainty.
Abstract
Flying ad-hoc networks (FANETs) are decentralized systems of autonomous unmanned aerial vehicles (UAVs) that must self-organize and sustain connectivity in dynamic and uncertain environments. These networks are invaluable where infrastructure is absent or damaged, yet they face strict energy limitations: battery-powered UAVs acting as communication relays near gateways deplete faster, leading to the hot-spot problem and progressive network fragmentation. This paper introduces RAMNA, a resource-aware optimizing algorithm designed to maximize network availability by autonomously repositioning UAVs at runtime. RAMNA continuously monitors network energy balance and link quality, and triggers adaptive role-swaps among UAVs to prevent premature depletion of critical nodes. We evaluate RAMNA across 66 heterogeneous topology scenarios, comparing it with the state-of-the-art Swap Level algorithm and a passive baseline, using real UAV energy consumption profiles that include both mobility and communication costs. Results show that RAMNA increases network availability by 85.5%–175.9% over passive operation and achieves up to 23.8 percentage-point improvement over Swap Level. We discuss how RAMNA contributes to the self-healing and self-optimization properties required for resilient, long-lived flying ad-hoc networks operating under energy uncertainty.