Abstract

In Recent years, Mobile Wireless Sensor Networks (MWSNs) have become fairly widespread because they can offer flexible and timely monitoring in changing situations like disaster management, environmental observations, healthcare, and military surveillance. Yet, a big challenge is the limited energy of mobile sensor nodes, which affects how well the network performs, its stability, and its life span. Common routing methods often struggle to keep up with the constant changes in network structure and the unpredictable movement of nodes found in MWSNs. To address these issues, this paper introduces an Intelligent Energy-Aware Routing Protocol using Artificial Intelligence (IEARP- AI). This model focuses on using energy wisely while improving routing in mobile sensor networks. The proposed protocol integrates with machine learning method Fuzzy–PSO which is used to selects CHs by estimating the residual energy, distance to sink, node density, and mobility factor. Then the routing using an Reinforcement Learning algorithm called Deep Q-Learning (DQL) method which is continuously learns and updates optimal routing paths based on reward functions like energy usage, delay, and link stability. By combining AI-driven optimization and clustering techniques, IEARP-AI minimizes redundant data transmission, balances energy utilization among nodes, and extends the overall network lifetime. The Simulation results shows that IEARP-AI outperforms conventional routing protocols in terms of Network Lifetime, Average Energy Consumption, Packet Delivery Ratio, End-to-End Delay, Throughput, Scalability Index, Computational Efficiency, Energy Efficiency, Robustness Index, Adaptability Rate and Generalization Error. This intelligent approach provides a promising solution for the development of sustainable and high-performance mobile wireless sensor networks.

Authors

R.U. Anitha
Sona College of Arts and Science, India

Keywords

Mobile Wireless Sensor Networks (MWSNs), Intelligent Routing Protocol, Energy Efficiency, Fuzzy-PSO Clustering, Deep Q-Learning, Reinforcement Learning, Node Mobility