Research On The Application Of Improved Dynamic Clonal Selection Algorithm In Intrusion Detection

With the rapid development of computer and communication technologies,computer appolication in kinds of industry and research fields are increasingly developing in breadth and depth,thus making computer system involved in intrusion problem. Thus Network security has become more remarkable and complicated,making computer security involved in business secret,personal benefits and national security.Network security is not only a safety problems, but also a global important information security problem.There are several techniques mainly used in network security,such as firewall,security router and identity certification system.These techniques belong to category of static security technologies which defend against Unauthorized intrusion in some way.But only defense is quite unsatisfied in security management.Contrast to static security technology,dynamic strategy can meet the requirement.Intrusion detection technology is a kind of dynamic strategy,which Can real-timely monitor the network and dynamically protect the network when suffering from attack.Immune system and the Intrusion Detection System are wonderful similar.The former protects bodies from kinds of pathogen such as bacteria, viruses, etc, while the latter protects the computer system from risk or threat of intrusion more or less.Both of them are maintaining the stability of the system in constantly changing environment. Naturally,there is a real meaning and broad prospects that introduces the immunology principles and methods to intrusions detecting fields.In the first place,this thesis introduces knowledge of network security and the gloss,function,modeling methods and classification of the IDS.A brief introduction about the intrusion detection technology is made,and we conclude their drawbacks,point out the developmental direction in which the IDS will follow in future.In the second place, we research some basic theories of immunology, and introduces the basic concept of artificial immune system(AIS) and its future,the immune algorithms and some applications of AIS on IDS.Further more,as the basic algorithm of immunity algorithm, the Clonal Selection algorithm has received very many of attention.Based on the research immunity principle and in the clone selective principle foundation,this article has made some improvement in the existing Clonal Selection algorithm, which adds the gene bank and the high controllably variation to the dynamic clone choice algorithm,enhancing the detector the evolution speed.At the same time , the variation and the competition are added to the memory detector production,which promoting memory detector evolution and enhancing the affinity level of population.In the modeling,we validate the system prototype , by using the network invasion test collection that provided by invasion research group of Computer Science Department in American Columbia University. The system simulation experiment result had proved that,the algorithm after improvement has made some enhancement in the convergence rate and the examination efficiency.