改进的基于人工免疫的入侵检测模型

AN IMPROVED ARTIFICIAL IMMUNE-BASED INTRUSION DETECTION MODEL

针对现有的基于人工免疫的网络入侵检测系统存在生成检测器效率不高,且记忆检测器无法很好地适应动态变化的网络环境等缺陷,在Kim小组提出的动态克隆选择算法DynamiCS的基础上进行改进,提出新型的网络入侵检测模型。该模型在基因库生成检测器的算法上进行改进,设计有效的基因变异重组算法,以期高效地产生更多的合格检测器;设计并采用改进的记忆检测器更新算法,以保证记忆检测器的活性。最后,对新模型进行了网络入侵检测仿真实验,验证了所提模型的可行性和有效性。

Aiming at the limitations of existing artificial immune-based network intrusion detection system that it is low efficient in generating the detector and its memory detector is not able to well adapt to the networks environment with dynamic variation, we propose a novel network intrusion detection model based on the improvement of dynamic clonal selection algorithm DynamiCS presented by Kim＇ s team. The model improves the algorithm of detector generation in gene bank and designs effective gene mutation recombinant algorithm in order to generate more qualified detectors efficiently. We design and adopt the modified memory detector updating algorithm to guarantee the activity of the memory detectors. At last, the simulation experiment of network intrusion detection is made on the new model, it verifies the feasibility and the effectiveness of the proposed model.