No abstract

Abstract. The method of intrusion detection of embedded network equipment in large assoc iation is researched. In the process of large-scale embedded network device associated intrus ion detection, the intrusion detection results of network equipment directly affect the stabilit y and security of the network. For this, an intrusion detection method of large-scale embed ded network device is proposed based on improved ART2. When there are amount of mem ory models in artificial neural network, effective organization for learning the model can be carried out, and improve the detection efficiency, the judgment condition adjustment is red uced by linear combination of amplitude and phase, the cluster size difference is reduced, t hus, the network intrusion detection and positioning of device are completed. The experimen t results show that, by using the improved ART2 algorithm for large embedded network de vice intrusion detection, it can simplify the training set, shorten the detection time, the accu racy of detection is improved.