Enhancing network longevity in WSNs via a two-layer hierarchical routing protocol with dual-hexagonal topology.

Wireless Sensor Networks (WSNs) face a critical challenge in energy efficiency and reliable data transmission. The Clustering of networks plays an important role in enhancing network longevity. The latest innovations achieve the advantages of hexagonal topologies over traditional square grids in terms of coverage and connectivity. An innovative two-layer hierarchical routing protocol centered on a dual hexagonal architecture called as DS-HHP, DQN-SARSA Hybrid Hexagonal Routing Protocol for WSN, is presented in this paper. The network is first divided into hexagonal clusters. Sensor nodes are then organized in a secondary hexagonal pattern, and the Cluster Head (CH) is located at the center of each cluster. The proposed DS-HHP design integrates a dual-hexagonal deployment structure with a two-layer reinforcement learning strategy: a Deep Q-Network (DQN) for strategic Cluster Head (CH) selection and a SARSA-based routing agent for adaptive intra-cluster communication. Key learning parameters, including learning rate (α), discount factor (γ), and exploration rate (ε), are carefully tuned to ensure stable convergence and efficient decision-making. Simulation results demonstrate that DS-HHP significantly outperforms benchmark schemes, including ERGR-EMHC, EEGT, and VHFRP. The DS-HHP establishes a strong foundation for intelligent and self-optimizing WSNs by combining structured network topology with multi-level reinforcement learning.