Abstract: ID model-IAIDM(Immune-based Adaptive Intrusion Detection Model) is put forward. Because current IDS usually has a problem of incomplete training sets, this page has done the deep researches on the problem caused by the limit of training time with the feature of IAIDM. Based on it, an incremental algorithm(IA) is put forward. The experiment results demonstrate IA can update local selfspace that has changed incrementally and dynamically instead of the whole space so that IAIDM can adjust itself to the current network environment quickly.

Key words: Intrusion detection, Natural immune system, Incomplete

摘要： 提出了一种基于免疫原理的自适应入侵检测模型IAIDM，由于训练集非完备性问题是当前入侵检测系统遭遇到的最常见的问题，因此该文结合IAIDM模型特点，对因时间因素而导致训练集非完备性问题进行了深入分析，提出了增量式动态更新算法IA，实验结果显示IA能增量式地动态更新发生变化的局部样本空间而不必更新整个样本空间，保证了IAIDM能迅速适应网络环境的变化。

关键词: 入侵检测, 自然免疫系统, 非完备性