A multicriteria model for prioritizing 5G network deployment with Monte Carlo stability analysis: A case study in Magdalena, Colombia

The global deployment of 5G mobile networks has followed a progressive and strategic approach, depending on regional characteristics. This study develops a reference framework to prioritize areas for deployment, considering technical, sociodemographic, geographic, and economic criteria. The methodology integrates multi-criteria decision-making techniques such as AHP, CRITIC, TOPSIS, and SAW to evaluate and prioritize alternatives. This framework was applied to the municipalities of the Magdalena department in Colombia. The AHP results indicate that technical criteria are the most relevant in the selection process, with a weight of 34.3%, followed by sociodemographic criteria at 33.6%, geographic criteria at 19.47%, and economic criteria at 12.63%. A high similarity in municipality prioritization was observed, with a correlation of ρ = 0.9897 according to Spearman's coefficient. Using TOPSIS and SAW, the municipality of Ciénaga ranks first, given that the sub-criteria of population size, area, and 4G coverage hold the highest relevance in the selection process, with percentages of 13.13%, 12.16%, and 11.12%, respectively, while the municipalities of Fundación and Plato alternate between second and third place. On the other hand, La Zona Bananera ranks fourth and fifth. To assess the model's robustness against variability in the criterion weights, a sensitivity analysis was conducted using the Monte Carlo method with 10,000 iterations. The results indicate that the ranking remains stable, with an average correlation of ρ = 0.9010 between the rankings obtained with SAW and the final ranking using TOPSIS. The influence of high-weight and highly correlated sub-criteria was also assessed. Rankings from AHP-SAW and AHP-TOPSIS were compared with CRITIC-SAW and CRITIC-TOPSIS, yielding correlations of 0.98 and 0.76, respectively. It can be concluded that the deployment of 5G networks can be systematically prioritized based on the mentioned criteria, using a model that remains stable despite changes in criterion weights.