The widely acceptable problem in wireless sensor networks (WSNs) is to develop a practical scheme for data aggregation in the massive range of sensor nodes that are randomly distributed over a network region. The essential operation of cluster heads (CHs) in such a network is to transmit the aggregated data to the sink node through multi-hop communication, thus the energy to be used in a better way during the period of aggregation and transmission. Therefore, this study presents a scheme based on grid clustering and fuzzy reinforcement-learning to maximise network lifetime as well as energy-efficient data aggregation for distributed WSN. Initially, grid clustering is employed for cluster formation and CH selection. Further, a fuzzy rule system-based reinforcement learning algorithm is used to select the data aggregator node based on the parameters, such as distance, neighbourhood overlap, and algebraic connectivity. Finally, the dynamic relocation of the mobile sink is performed within a grid-based clustered network region using a fruit fly optimisation algorithm. The experimental outcomes revealed that the proposed data aggregation scheme provides superior performance in terms of energy consumption and network lifetime compared to earlier systems.

中文翻译：

---

基于网格聚类和模糊强化学习的分布式无线传感器网络节能数据聚合方案

无线传感器网络（WSN）中广为接受的问题是开发一种实用的方案，用于在随机分布在网络区域中的大量传感器节点中进行数据聚合。在这种网络中，簇头（CH）的基本操作是通过多跳通信将聚合数据传输到宿节点，因此在聚合和传输期间将以更好的方式使用能量。因此，本研究提出了一种基于网格聚类和模糊增强学习的方案，以最大化网络寿命以及分布式WSN的节能数据聚合。最初，将网格聚类用于聚类形成和CH选择。此外，基于模糊规则系统的强化学习算法用于根据参数选择数据聚合器节点，例如距离，邻域重叠和代数连接。最后，使用果蝇优化算法在基于网格的群集网络区域内执行移动接收器的动态重定位。实验结果表明，与较早的系统相比，所提出的数据聚合方案在能耗和网络寿命方面提供了卓越的性能。