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.

无线传感器网络已成为物联网 (IoT) 应用的重要组成部分。由于其资源限制性质，实现 QoS 要求的重大挑战包括优化能源利用、延长寿命、最小延迟、足够的数据包传递率等。最近的文献中讨论了许多解决这些问题的优化和路由方法。但是，它们在处理具有复杂潜在分布的高维数据时存在局限性。因此，在本文中，我们提出了一种能量感知神经模糊路由模型（EANFR），它处理高能传感器节点以形成集群并在深度神经网络生成的特征空间中做出路由决策，以解决问题。经过训练的 EANFR 模型可以选择合适的簇头节点和最有活力的路由，最短路径。进行了系统全面的仿真，统计分析结果表明，所提出的EANFR模型获得了最低的训练误差。此外，EANFR 在网络寿命方面优于最近的文献，特别是在使用神经模糊方法的能量感知聚类方面，提高了 89.23%，自适应 Q 学习提高了 67.21%，径向基模糊神经网络 2 型模糊权重提高了 20.63%。根据这项研究，所提出的 EANFR 模型显着提高了 WSN 的网络寿命和 QoS 性能，使其适用于物联网监控应用。EANFR 在网络寿命方面优于最近的文献，特别是在使用神经模糊方法的能量感知聚类方面，提高了 89.23%，自适应 Q 学习提高了 67.21%，径向基模糊神经网络 2 型模糊权重提高了 20.63%。根据这项研究，所提出的 EANFR 模型显着提高了 WSN 的网络寿命和 QoS 性能，使其适用于物联网监控应用。EANFR 在网络寿命方面优于最近的文献，特别是在使用神经模糊方法的能量感知聚类方面，提高了 89.23%，自适应 Q 学习提高了 67.21%，径向基模糊神经网络 2 型模糊权重提高了 20.63%。根据这项研究，所提出的 EANFR 模型显着提高了 WSN 的网络寿命和 QoS 性能，使其适用于物联网监控应用。