Wireless sensor networks are used to track and regulate physical conditions like temperature and the environment's humidity. Wireless sensor networks with their advanced features are adopted for many real-time applications. The limited capacity batteries usually power the sensor nodes. But the big challenge of limited battery capacity obstructs remote and inaccessible areas where their use is the most favorable. For extending the lifetime of the network, the battery should optimally utilize it for different operations. The requirement toward low complexity and low energy consumption motivate the wireless sensor networks' efficient clustering algorithm. The sensor nodes group into clusters; one sensor node is chosen as a cluster head, and communication to the sink node from the sensor nodes occurs through the cluster head (CH) node. In the proposed method, cluster heads are determined based on the sensor node's weighted metric. The sensor nodes are then self-adaptive by making correct decisions in real-time based on the sensed data, but detected information is often inaccurate due to some mechanical, wireless loss, and battery problems. The erroneous or irrelevant data should be overlooked to avoid unnecessary data transmission, which contributes to reducing the network's lifetime. In the neighborhood-dependent self-diagnosis fault detection technique, the faulty sensed data are filtered at the sensor node itself. In the data prediction algorithm, the filtered data are predicted at the cluster head. All the factors collectively contribute to enhance the network's lifetime. The lifetime improvement of the proposed approach is almost doubled compared with LEACH one time better than QLEACH and ECH, and 51% better than temporal approach.

中文翻译：

---

一种具有自诊断数据故障检测和预测的无线传感器网络节能聚类算法

无线传感器网络用于跟踪和调节温度和环境湿度等物理条件。许多实时应用都采用了具有先进功能的无线传感器网络。容量有限的电池通常为传感器节点供电。但电池容量有限的巨大挑战阻碍了偏远和人迹罕至的地区，而这些地区最适合使用它们。为了延长网络的使用寿命，电池应针对不同的操作最佳地利用它。低复杂度和低能耗的要求激发了无线传感器网络的高效聚类算法。传感器节点分组为簇；选择一个传感器节点作为簇头，从传感器节点到汇聚节点的通信通过簇头（CH）节点发生。在所提出的方法中，簇头是根据传感器节点的加权度量确定的。然后传感器节点通过基于感测数据实时做出正确决策来自适应，但由于一些机械、无线损失和电池问题，检测到的信息通常不准确。应该忽略错误或不相关的数据，以避免不必要的数据传输，这有助于缩短网络的生命周期。在基于邻域的自诊断故障检测技术中，在传感器节点本身过滤有故障的感测数据。在数据预测算法中，过滤后的数据在簇头进行预测。所有因素共同有助于提高网络的寿命。