Análise comparativa de algoritmos meta-heurísticos na solução de posicionamento de UAV-BS

Abstract

In natural disaster scenarios, communication infrastructure is often compromised, hindering rescue and support operations. In this context, Unmanned Aerial Vehicles (UAVs), acting as Base Stations (UAV-BS), emerge as a promising alternative to restore connectivity. Although several studies investigate metaheuristic algorithms for optimizing UAV positioning, many rely on idealized conditions and overlook the random distribution of users typical of emergency situations. This study presents a comparative analysis of five metaheuristic algorithms: Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Artificial Bee Colony (ABC), Ant Colony Optimization (ACO), and Teaching-Learning-Based Optimization (TLBO). The comparison considers key performance metrics such as user coverage, throughput, execution time, and signal quality. Statistical tests revealed significant differences among the algorithms. PSO and TLBO achieved the best overall performance, while ACO delivered the weakest results in terms of coverage and execution time. TLBO proved to be the most balanced approach, whereas PSO stood out in scenarios that prioritize reducing the number of UAVs. The findings provide valuable insights to support more effective decision-making in emergency mobile network deployments.