Research On Improved NA-based Anomaly Detection Model

The researchers pay more attention on using the mechanism of biological immune system for reference to solve the net security problem. Especially, the field of amusedly-based detection becomes a hot research field.The author researches the basic theory and principle of artificial immune and comes up with an improved NA-based Anomaly Detection.The main research is as follows.First,analyzing the present security technology and security system theory, summarizing the past and present of the intrusion detection technology, analyzing the shortcoming of the present intrusion detection in detail,and pointing the IDS research direction.Second,researching the negative algorithm,analyzing the detection capacity of this algorithm,pointing the potential holes: length-limited holes and crossover holes,discussing the cause for the holes and how to reduce the hole's number.A further study on how to address the length-limited holes problem.The auther compares the r-continuous matching rule with the r-chunk matching rule and points out that the r-chunk matching rule makes up the length-limited holes which caused by r-continuous matching rule. Meanwhile, the auther comes up with two generator B and T to recognize the abnormity.Last, making a test on the hole problem and the improved NA-based Anomaly Detection.