Research On Detector Generating Algorithm And Analyzer Model In Immunity Intrusion Detection

Along with the rapid development of network technology, the networkenvironment and attack methods of hackers become more and more complex. Thetraditional intrusion detection can not adapt to this ever-changing network.TheArtificial Immune System (AIS) is a new computation theory inspiredbybiological immune system. The good characteristic of AIS withdistributivity,robustness and self-organization makes the Intrusion DetectionSystem (IDS)based on AIS been a hot spot in the research of network security.From the perspective of improving the recognition and distributionperformance of detector, a real-valued negative selection algorithm based on PCAis proposed inspired by performance improving mechanism of antibody cell inbiological immune system. This algorithm uses principal components extractedby PCA to construct the low dimensional shape space for improving therecognition ability, and employs weighted Euclidean distance as the matchingrule to training detectors in principal component shape space for improving theability of detector. The experiments compare this algorithm with traditionalreal-valued negative selection algorithm on KDD CUP99datasets. Experimentalresults show that this algorithm can supply the deficiency of real-valued detectorgeneration algorithm in high dimensional space, and improve the detectionperformance of detector in high dimensional space.Aiming at the defects in unknown intrusion detection, the detector don’tupdate timely and other issues, I propose an improved immunity intrusiondetection analyzer model. The immunity immune intrusion detection modelcombines the random generation and real-valued negative selection algorithmbased on PCA to generate diverse and efficient mature detector,and designeffective memory detector updating algorithm to improve the effectiveness ofmemory detector.