The article addresses the challenge of efficient information transmission in wireless sensor networks (WSNs), focusing on enhancing network lifetime through energy-aware clustering and cluster head (CH) selection. Traditional methods often underperform under varying environmental conditions and uncertainties. To overcome this, the authors propose a novel approach combining fuzzy reinforcement learning with meta-heuristic optimization algorithms—Harris Hawk Optimization (HHO) for energy-aware clustering and Wild Horse Optimization (WHO) for training fuzzy rules in the reinforcement learning system. This hybrid system adapts dynamically to environmental changes, optimizing cluster formation based on node energy, location, and distribution, and selecting CHs by considering residual energy, node neighborhood density, and distance to the base station.

The fuzzy logic technique plays a central role by handling uncertainties and imprecision inherent in WSNs, using membership functions and if-then rules to prioritize nodes for CH selection. The WHO algorithm optimizes the fuzzy rule set, improving decision-making in the reinforcement learning framework. The HHO algorithm clusters nodes by minimizing intra-cluster variance and balancing energy distribution. The integrated fuzzy reinforcement learning system, trained via WHO, effectively selects optimal CHs, reducing power consumption and extending network lifetime. The approach is validated through objective functions and simulations demonstrating improved clustering and CH selection performance, highlighting the system’s adaptability and robustness in real-world WSN deployments.