Metaheuristic approaches for optimizing UAV base station placement in THz communication for disaster recovery scenarios


Ersin I., Demirci S., Akkas M. A., DAĞDEVİREN O.

Ad Hoc Networks, vol.194, 2027 (SCI-Expanded, Scopus)

  • Publication Type: Article / Article
  • Volume: 194
  • Publication Date: 2027
  • Doi Number: 10.1016/j.adhoc.2026.104386
  • Journal Name: Ad Hoc Networks
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Applied Science & Technology Source, Compendex, INSPEC
  • Keywords: Ad hoc networks, Base station placement problem, Metaheuristic algorithms, THz band, UAV communication
  • Ondokuz Mayıs University Affiliated: Yes

Abstract

Emergency communication networks are critical for disaster response and recovery operations. This study investigates the deployment of Unmanned Aerial Vehicle (UAV)-mounted base stations operating in the Terahertz (THz) band to establish rapid communication infrastructure in disaster-affected regions. We address the optimal UAV placement problem using seven advanced metaheuristic optimization algorithms: Particle Swarm Optimization (PSO), Genetic Algorithm (GA), Artificial Bee Colony (ABC), Firefly Algorithm (FA), Whale Optimization Algorithm (WOA), Grasshopper Optimization Algorithm (GOA), and Harris Hawks Optimization (HHO). Our comprehensive evaluation encompasses two operational scenarios: 120-second deployment windows and 500-iteration optimization cycles, across multiple altitude configurations (10m, 20 m, 30 m, 40 m) and UAV quantities (4, 6, 8 units). We analyze both throughput capacity (Tbps) and energy efficiency metrics, including energy consumption per bit (nJ/bit) and energy efficiency (bit/J), with a fixed power budget of 450 Watts per UAV. To address disaster-response feasibility, the evaluation further includes reliability-oriented outage metrics, LBLRTM/ITU-R-based adverse-weather sensitivity, and aperture-dependent pointing-error Monte Carlo analysis. A same-aperture THz/mmWave pointing-error comparison is further included to quantify whether the THz hotspot layer preserves its capacity advantage despite being more sensitive to beam misalignment. Results demonstrate that PSO and GA consistently outperform other algorithms, achieving energy efficiency values of (Formula presented) bit/J (120-second scenario) and (Formula presented) bit/J (500-iteration scenario) at 10 m altitude. The 500-iteration scenario exhibits approximately 2.5–3.5 times higher energy efficiency compared to the 120-second scenario. Energy per bit values range from 0.67 nJ/bit to 6.43 nJ/bit, all within acceptable industry standards for optical and THz communication systems. Increasing UAV quantity from 4 to 8 units improves total system capacity by approximately 2.0–2.2 times, though individual UAV efficiency decreases due to increased system complexity and interference. Our findings indicate that the optimal configuration combines the PSO algorithm, 500-iteration optimization cycles, and 6–8 UAVs at a 10 m altitude, delivering superior energy efficiency and capacity for emergency communication networks.