Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Computer network > General issues > Computer Network Security

Research of Immune Theory Application in Distributed Intrusion Detection System

Author YaoXing
Tutor GuoFan
School Jiangxi Normal University
Course Computer Software and Theory
Keywords Immune Principle Intrusion Detection Negative selection Clonal selection r consecutive matches
CLC TP393.08
Type Master's thesis
Year 2010
Downloads 30
Quotes 0
Download Dissertation

With the increasing size of the network , the network structure is very complex, the existing intrusion detection system is difficult to adapt to large-scale distributed network security needs. The immune system with its information processing in a distributed demonstrated protection, adaptability , robustness, scalability and memory ability and other characteristics, well adapted to the development needs of the network . Therefore immune theory in intrusion detection research is very necessary . Based on the relevant principles of biological immune system , we propose a new model for intrusion detection system architecture and design and implementation of an immune algorithm, experimental results , the algorithm can effectively detect abnormal data . This paper is designed based on immune principle Distributed intrusion detection model , the model based on the characteristics of the distributed network , the entire detection system is divided into a hierarchy Host IDS (HIDS), IDS central server , IDS and the IDS partition server root server and other components, immune theory used in the function of each component in the implementation . The entire model anomaly detection and misuse detection organically linked together and give full consideration to host the actual network environment faced by the host of the IDS detection process to simplify and improve the detection efficiency. Secondly, based on the realization of different immunological characteristics related algorithms, and combine them into a complete intrusion detection method . This method is called by negative selection algorithm , clonal selection algorithm and r consecutive matching algorithm components. To reduce the \Further, since the training data set generally is first randomly generated binary string of specified length , with its tolerance of negative selection algorithm and then matched with the exception of data sets , a successful match detector set to join . Because the way the randomness caused by the training process takes a very long time. In order to shorten the training time , this paper uses a fast training methods . In this paper, MIT Lincoln Laboratory KDDCup99 network intrusion detection data sets as experimental data set experiments . Experimental results show that the method of fast training data set can effectively reduce the time of the training data and the matching variable r further described algorithm can effectively reduce the \

Related Dissertations
More Dissertations