An Energy-Efficient Scheduling and Routing Protocol Based on Q-Learning for WSN

Purpose: This article introduces an energy-efficient routing protocol for wireless sensor networks (WSNs) that integrates a dynamic K-means clustering algorithm with Q-Learning and adaptive sleep scheduling. The proposed model aims to extend the network’s lifetime, reduce energy consumption, and maintain the reliability of high data delivery in limited-resource nodes. Methodology: Each sensor tag autonomously makes optimal forwarding decisions based on local parameters such as remaining energy, distance, jumping, link quality and sensory data variation. To increase adaptation, the network regularly prepares the cluster depending on the node energy and position. In contrast, the sensor nodes enter sleep mode when no significant data changes are detected, reducing inactive communication. Findings: The model was evaluated with separate network density and simulation settings in 25 scenarios. The best executive landscape achieved a package delivery ratio (PDR) of 94.73 %, delayed 5080.1 episodes in First Node Death (FND), and reduced the average energy consumption by up to 0.0111189 J per episode. Unique Contribution to Theory, Practice, and Policy: Compared to standard protocols such as LEACH and RLBEEP, the proposed method outperforms them in all performance matrices. These results demonstrated the effectiveness of learning combined with adaptive grouping and transmission control for achieving durable and intelligent WSN operation.