一种基于神经网络的黑客入侵检测新方法

A New Method for Internet Intrusion Detection Based on Neural Netoworks

给出了一个基于神经网络的网络入侵检测系统模型 .该模型可对网络中的 IP数据包进行分析处理以及特征提取 ,并采用智能神经网络进行学习或判别 ,以达到对未知数据包进行检测的目的 .首先建立功能专一、结构简单、易于构造的神经网络来完成单一的黑客入侵检测任务 ,然后利用智能神经网络组成原理将这些能够检测多种多样的黑客入侵的小网合并 ,组合成功能完善、结构复杂的大网来完成整个检测任务 .

A intrusion detection system based on intelligent neural networks has been presented. The system works by capturing packets and using a intelligent neural network to learn or identify an intrusive behavior within the analyzed data stream. A complex network which based on several single ones can be used to identify many kinds of internet intrusion. Experiment shows this is a good method for intrusion detection.