RBF神经网络抗体群用于入侵检测的研究

摘要

为了进一步模仿生物免疫系统,提升入侵检测系统的性能,提出了具有抗原针对性的抗体生成算法和具有抗体针对性的抗原检测算法.对不同类型的已知抗原,生成相应类型的抗体,对不同类型的外来抗原,用相应类型的抗体检测,提高了对生物免疫系统的模仿程度；通过免疫应答、克隆选择和疫苗注射来训练生成神经网络抗体群.仿真实验结果表明,该算法具有很高的精度和很强的自适应能力.%To further mimic biological immune system, improve the performance of intrusion detection system, a antigen specific antibody production algorithm and a antibody specific antigen detection algorithm are proposed. Known for different types of antigens to generate appropriate types of antibodies, different types of foreign antigens, with corresponding types of antibodies detection of biological immune system, the degree of imitation is improved. The neural network antibody group is generated through immune response, clonal selection and vaccination. Simulation results show that the algorithm has good precision and strong adaptability.