Energy-Efficient and Equitable Load Balancing Using Q-Networks (E3lb-Qnet) For Energy Aware Load Balancing in Manet

Abstract

In mobile ad-hoc networks (MANETs), efficient load balancing and energy management are critical to enhancing network performance and extending its operational lifespan. This paper proposes Energy-Efficient and Equitable Load Balancing using Q-Networks (E3LB-QNet), a novel approach leveraging Deep Q-Networks (DQN) to dynamically balance traffic loads across nodes while considering energy constraints. The algorithm incorporates real-time feedback from network conditions such as node load, residual energy, and link quality, enabling nodes to make intelligent, energy-aware routing decisions. By utilizing reinforcement learning, E3LB-QNet ensures equitable load distribution, prevents overloading of energy-constrained nodes, and extends the overall network lifetime. Experimental results demonstrate that E3LB-QNet significantly improves energy efficiency and load balancing compared to traditional methods, leading to enhanced throughput, reduced packet loss, and prolonged network longevity.