A Terrain- and SNR-Aware Framework for UAV-FSO Backbone Deployment in Post-Disaster Networks: AeroFSO-Deploy

Abstract

Natural catastrophes frequently result in the partial or complete destruction of terrestrial communication networks, which severely degrading command, control, and coordination capabilities. Unmanned Aerial Vehicles (UAVs) equipped with free-space optical (FSO) transceivers offer a promising solution for a rapidly deployable, high-throughput, and interference-free backbone. However, designing these aerial networks remains challenging, as FSO links require strict line-of-sight (LoS) and sufficient signal-to-noise ratio (SNR), while UAV positions must ensure both ground user coverage and global network connectivity over complex terrain. In this paper we propose AeroFSO-Deploy, a modular, terrain aware deployment framework for UAV-FSO backbones in post-disaster scenarios. The method combines (i) a relaxed set-cover formulation for coverage, (ii) an SNR-validated connectivity restoration phase operating on a DEM-pruned graph, and (iii) an iterative pruning stage that removes redundant UAVs without violating coverage or connectivity. Numerical experiments on three realistic digital-elevationmodel (DEM) terrains show that AeroFSO-Deploy achieves full coverage and LoS+SNR-validated connectivity with 22% to 35% fewer UAVs than greedy, random, and connected-dominating-set baselines. With runtimes under 5 seconds on a standard desktop. These results demonstrate that AeroFSO-Deploy is well suited for fast, physically consistent network plan ning in time-critical disaster response.