Energy-aware neuro-fuzzy routing model for WSN based-IoT

Wireless sensor networks have become a vital part of the Internet of Things (IoT) applications. Due to its resource constraints nature, significant challenges in achieving QoS requirements include optimal energy utilization, enhanced lifespan, minimum delay, adequate packet delivery ratio, etc. Many optimizations and routing methods to solve these issues have been discussed in recent literature. However, they have limitations when dealing with high-dimensional data with complex latent distributions. Thus, In this article, we propose an energy-aware neuro-fuzzy routing model (EANFR) that deals with the high-energy sensor nodes to form the clusters and make routing decisions in a feature space generated by a deep neural network to solve the problem. The trained EANFR model can select appropriate cluster head nodes and routes over the most energized, shortest path. A systematic and comprehensive simulation was carried out, and the statistical analysis results show that the proposed EANFR model acquired the lowest training errors. Furthermore, the EANFR outperforms recent literature in terms of network lifetime, particularly on energy-aware clustering using neuro-fuzzy approach by 89.23%, Adaptive Q Learning by 67.21%, and Radial Basis Fuzzy Neural Network Type 2 Fuzzy Weights by 20.63%. According to this research study, the proposed EANFR model significantly improves the network lifespan and QoS performances of WSN making it suitable for IoT monitoring applications.