中文摘要：

结合入侵检测的实际需求，提出了一种集成人工免疫系统（IAIS）。该系统结合了树突状细胞算法（DCA）和否定选择算法（NSA），DCA用于检测行为特征，NSA用于检测结构特征。通过KDD99数据集实验对该系统进行验证，并与其他方法进行了比较。实验结果表明，IAIS检测性能与经典分类算法相当。IAIS具有不依赖明确标识的数据来训练检测器，可结合行为特征和结构特征进行实时入侵检测的特点。

英文摘要：

According to the practical requirements of intrusion detection, an integrated artificial immune system （IAIS） was proposed. The system combined dendritic cell algorithm （DCA） and negative selection algorithm （NSA）. DCA was used to detect behavioral features. NSA was used to detect structural features. IAIS was validated on KDD 99 dataset. Comparisons to other approaches were made. The experimental results show that the detection performance of IAIS is comparable to classic classification algorithm. IAIS does not rely on labeled data to train detectors. It combines behav- ioral features and structural features to detect intrusions in real-time mode.