Find in Library

In order to solve the problems of low detection accuracy and less attention of vairous malicious traffic identification in traditional intrusion detection system (IDS) in the face of massive data mixed with various malicious traffic, an intrusion detection model based on capsule network CapIDS was proposed. In this model, the traffic to be measured was input into the capsule network in the form of capsules, and the characteristics of malicious traffic were extracted by using the dynamic routing mechanism to complete the identification of malicious traffic. At the same time, the structure of the capsule network was improved, which made the model have stronger generalization ability and improved the detection accuracy of malicious traffic. The experimental results show that the detection accuracy of the proposed model for malicious traffic has reached 99.50% on the NSL-KDD data set, which is better than that of other schemes.