Optimization Method Of Immunity Intrusion Detector Based On Rough-set And Crowding-Niche

With the rapid development of network,network security risks are correspondingly more and more.So the importance of network security protection is self-evident,and then a variety of network security technologies have emerged.Because of its aggressiveness,the intrusion detection system has become an important means to ensure network security.Biological immune system has many advantages,such as dynamic and adaptive.It has many similarities with intrusion detection system,and the application of immune mechanism in intrusion detection system has become a hot topic.However,for the intrusion detection system based on immune mechanism,the quality of the detector is the key to determining the performance of the system.Therefore,the optimization of detectors is favored by many scholars,and becomes one of the most important research in the field of intrusion detection.At first,this paper analyzes the problems of the initial generation of the immune intrusion detector in the real-value space and combines the rough set theory to propose an optimization algorithm of detector's generation based on the rough set.using the rough set reduction theory to transform the high dimensional shape-space into low dimensional shape-space on the premise of ensuring that the information classification ability is not affected,and using weighted Euclidean distance by attributes significance of rough set to calculate the affinity degree toadjust detectors,so that the detectors can cover the non-self set better and not cover the self set.The experimental results show that the optimized detector can greatly reduce the overlap rate between the detectors and improve the detection performance of the detector under the premise of ensuring a certain coverage rate.This paper then analyzes the problems of basic immune clonal selection algorithm in detector evolution stage,and combines niche technology to put forward an improvement idea.in the process of clonal selection,using crowding-niche technology to mature detector Evolution,and encouraging the detectors of different populations to hybridization,mutation.In addition,using the rough set theory to generate the initial group of this algorithm.Experiments show that the optimized algorithm increases the diversity of the detectors,thus improving the detection rate of unknown intrusion.