基于改进BP神经网络的智能Agent分布式入侵检测系统

INTELLIGENT-AGENT DISTRIBUTED INTRUSION DETECTION SYSTEM BASED ON IMPROVED BP NEURAL NETWORK

针对目前入侵检测技术存在问题。根据通用入侵检测框架CIDF,给出了一个基于改进BP神经网络的多Agent分布式入侵检测模型MAIDMBN(Multi-Agent Distributed Intrusion Detection Model Based on Improved BP Neural Network),该模型采用了异常检测与误用检测相结合和基改进BP算法的学习机制。MAIDMBN的实验结果表明在误报率、漏报率有一定的改善,系统能进行有效的检测。

Aiming at the limitation of the existing technologies to the intrusion detection system. Based on common Intrusion Detection System Framework CIDF. In this paper, based On improved Neural Network Agent Distributed intrusion Detection System, The model use of the anomaly detection and misuse detection and combining algorithm based on Improved BP learning mechanism. MAIDMBN experimental results show that the rate of false alarm and the rate of failing to report are improved , the system can be effectively detected.