An Energy Efficient Routing for WSN Using Adaptive Opposition Based Reinforcement Learning

The proposed study contains four stages: Network Setup phase, Clustering, Data Aggregation, and optimal routing. At first, the sensor nodes are located, and the energy model is initialized in the WSN. Then, the sensor nodes are grouped into a cluster using kernel based Fuzzy C means clustering (K-FCM). This clustering process can increase the network lifetime (NL) in WSN. Next, the cluster head (CH) is chosen from each cluster group using the adaptive Elephant herd optimization (AEHO) approach. This approach chooses the CH based on parameters like residual energy, the distance between node and CH, the distance between CH and BS, and node degree. Finally, the selected CH aggregates the data using the adaptive reinforcement learning model. Here, the reinforcement learning algorithm is utilized for data aggregation. An opposition-based learning arithmetic optimization algorithm (OBL-AOA) approach is employed to route the aggregated data to the BS. The experimental analysis portrays that the proposed model attains enhanced performance compared to other existing approaches.