基于累积和控制图的分布式传感网络故障诊断

由于无线气象传感网具有资源受限及分布式等特点，传感器节点的故障诊断面临着很大挑战。针对现有诊断方法误报率高、计算冗余量大的问题，提出了一种基于累积和控制图（CUSUM）与邻居协作融合的故障诊断方法。首先，通过累积和控制图分析传感器节点上的历史数据，提高对节点故障判断的灵敏度并且定位出异常时间点；然后，结合网络内邻居节点间的数据交换，通过判断节点的状态诊断出故障节点。实验结果表明，即使在整个网络中在节点故障率高达35%时，算法检测精度仍然高于97.7%，而误报率不超过2%。由此可见，在节点故障概率很高的情况下，此所提法也能得到很高的检测精度和较低的误报率，受节点故障率的影响明显减小。

Distributed fault detection for wireless sensor network based on cumulative sum control chart

With the stringent resources and distributed nature in wireless sensor networks, fault diagnosis of sensor nodes faces great challenges. In order to solve the problem that the existing approaches of diagnosing sensor networks have high false alarm ratio and considerable computation redundancy on nodes, a new fault detection mechanism based on Cumulative Sum Chart (CUSUM) and neighbor-coordination was proposed. Firstly, the historical data on a single node were analyzed by CUSUM to improve the sensitivity of fault diagnosis and locate the change point. Then, the fault nodes were detected though judging the status of nodes by the data exchange between neighbor nodes. The experimental results show that the detection accuracy is over 97.7% and the false alarm ratio is below 2% when the sensor fault probability in wireless sensor networks is up to 35%. Hence, the proposed algorithm has a high detection accuracy and low false alarm ratio even in the conditions of high fault probabilities and reduces the influence of sensor fault probability clearly.