Research on abnormal node detection in a wireless sensor network based on random matrix theory

Because the traditional detection methods have the problems of low recall and precision and long detection time, this paper studies a method of abnormal node detection in a wireless sensor network (WSN) based on random matrix theory. This method uses particle swarm optimisation to improve DV-Hop, and uses the improved DV-Hop method to locate WSN nodes. According to the spatiotemporal characteristics of WSN data, a data matrix is built, and the dimensionality of the data matrix is reduced by using a random matrix. The node attributes are judged according to the element correlation between multiple matrices to realise abnormal node detection in a WSN. The test results show that the average recall rate and recall rate of this method are 97.0% and 97.2% respectively, and the detection time is always less than 0.5s, so the practical application effect is good.