Research On Intrusion Detection Methods For The Internet Of Thing

Since the International Telecommunication Union(ITU)formally proposed the concept of the Internet of Things(Io T)in 2005,it has quickly gained recognition worldwide and become one of the core supports of the third wave of information industry revolution and the fourth industrial revolution.While the Io T industry continues to grow rapidly,security issues gradually become one of the obstacles that are limiting its extensive deployment.This thesis therefore conducted research on the following topics.(1)For the detection problem of node capture attacks in wireless sensor networks(WSN),we analyzed the characteristics of sensor nodes when they are captured,established a sensor node hierarchical network topology model,and proposed a node capture detection method using heartbeat sequence values.Comparative simulations were conducted with two other methods,and the results showed that our proposed method outperforms the compared ones.(2)For intrusion detection problems in the Io T network,first,a classifier model based on deep learning was constructed.Then,the data used were preprocessed,and finally,classification experiments were performed using the established model.Multiple indicators were used to evaluate the classification effect of the model.In addition,we compared the model we built with traditional machine learning algorithms and other classifier models.The experimental results showed that the classification model we built has good performance.(3)For practical application research on intrusion detection methods,we first introduced the Io T platform built by the subject group and enterprise cooperation projects,analyzed the security requirements of the platform,and proposed corresponding solutions.Then,by simulating cyber attacks on Io T gateway and collecting relevant network traffic data,the preprocessed data were classified using the classifier we built.Based on the evaluation indicator results,our work achieved certain results.Finally,we successfully applied and extended the intrusion detection subsystem based on the existing Io T platform.Example verification showed that this provides a practical solution for safeguarding Io T security.