Due to frequent changes in network topology,nodes in wireless sensor networks(WSNs)may move or fail,resulting in uncer-tainty in the data generated by sensor nodes.Node anomalies also exhibit multimodal characteristics,which can easily lead to issues of incorrect or missed detections.Therefore,a WSN anomaly node localization method that integrates decision threshold and trust filtering mechanism is proposed.A K-nearest neighbor graph signal function based on sensor location features is built to capture spatial relation-ships between nodes.A decision threshold based on the difference in smoothness before and after low-pass filtering is design to reduce the impact of data uncertainty and achieve WSN anomaly node detection.By using Beta distribution to preliminarily evaluate anchor point location information,adjusting trust update weights to adapt to topology changes,and introducing trust filtering mechanisms to dif-ferentiate WSN nodes,the credibility of anchor points is determined by cluster head nodes,achieving accurate localization of WSN ab-normal nodes.The results show that the proposed method has a false alarm probability of less than 1%for abnormal node localization,and the OPR value can reach 99.12%,indicating high localization accuracy.
万方数据
维普
