Enhancing the longevity and scalability of wireless sensor networks (WSNs) is achieved through the optimization of communication distances. To achieve this, a hierarchical clustering algorithm is employed, which helps conserve energy by facilitating efficient data collection and transmission among nodes. While previous clustering methods predominantly utilized a two-level hierarchy, limited attention has been given to a three-level framework. In our study, we introduce a clustering approach based on a three-level hierarchy, incorporating both centralized and distributed algorithms for the selection of upper-level and lower-level heads. The simulation results suggest that this novel technique performs effectively in improving network performance.