A Hybrid MO-ACO and Deep Reinforcement Learning Framework for Energy-Aware Routing in Wireless Sensor Networks

Wireless Sensor Networks (WSNs) have not yet overcome the vital issues of energy conservation and adaptive communication. The research study proposes MOACO-DQN, a hybrid energy-aware framework that utilizes Multi-Objective Ant Colony Optimisation (MO-ACO) and Deep Q-Network (DQN) for the intelligent management of networks to address these problems. The proposed method optimizes the selection of Cluster Heads (CHs) by MO-ACO, considering the residual energy, intra-cluster distance, and the CHs-sink distance. This ensures that energy is evenly used across the network. Using this framework, the proposed mechanism uses network state parameters (energy of nodes, quality of links, distance to sink) to learn and choose the most suitable next hop intelligently. By incorporating reinforcement learning into the routing process, the framework enhances the energy efficiency of data transmission. The MOACO-DQN proposed makes the network more durable and offers a lower end-to-end delay and a larger packet delivery ratio. The Proposed MOACO-DQN outperforms existing systems when the number of nodes is set at 100, with improvements of 20%, 25%, 35% over EEDQN, RL-LEACH, and BLE ACO, respectively. The suggested model is adaptable and better suited to smart and IoT-based WSNs.