基于变异的检测器生成算法研究

Research on variation-based detectors generating algorithm

为提高基于免疫的网络入侵检测系统中检测器的生成效率,减小计算量。对Forrest的否定选择算法(NSA)进行改进,提出候选检测器集的生成不再采用随机方式,而通过对两个数据集(一是已有的合格检测器集,二是自我数据集)进行变异来产生,即利用部分已有的检测结果反馈生成成熟检测器。改进算法提高了候选检测器成为成熟检测器的比率,实验结果表明了算法的有效性。

To enhance the efficiency of the detector generation in immune-based network intrusion detection system and reduce computation, Forrest＇ s negative selection algorithm （NSA） is improved. The candidate detector sets no longer are used randomly generated, and through the two data sets （the first is already qualified detector sets, and the second is self-data sets） to produce variation, namely, partially uses already the examination result feedback production mature detector which has. This improvement algorithm have had that the ratio ofthe candidate detector to become mature detector is enhanced, the experimental results show the effectiveness of this algorithm.