针对免疫入侵检测中实值空间存在的问题,借助免疫细胞的表位组织形式和离散拓扑理论,提出一种新的形态空间表示法—–邻域表示法.该方法利用数据的集合特性,采用空间中互不相交的邻域表示自体/检测器,并设计匹配策略和检测算法.实验表明,邻域空间可以较好地弥补实值空间的缺陷,提高检测器生成效率,改善系统整体检测效果.
For the problem of shape-space in the immunity-based intrusion detection system(IIDS),by using the epitope constitute of immune cells and discrete topology theory,a new method,named neighborhood is proposed.Furthermore,the corresponding matching rules and detectors generating algorithm are given.Experimental results show that,the new shape-space can solve the defects of real-valued representation,increase the efficiency of detector generation effectively,and improve the system’s detection rate.