| 本文已被：浏览 6724次 下载 7175次 |
码上扫一扫！ |
|
|
| 基于Agent的分布式入侵检测系统模型 |
|
马恒太1,2, 蒋建春1,2, 陈伟锋1,2, 卿斯汉1,2
|
|
1.中国科学院软件研究所信息安全国家重点实验室,北京,100080;2.中国科学院软件研究所信息安全国家重点实验室,北 京,100080
|
|
| 摘要: |
| 提出了一个基于Agent的分布式入侵检测系统模型框架.该模型提供了基于网络和基于主机入 侵检测部件的接口,为不同Agent的相互协作提供了条件.在分布式环境中,按照系统和网络的 异常使用模式的不同特征和环境差异,可利用不同的Agent进行检测,各Agent相互协作,检测 异常行为.该模型是一个开放的系统模型,具有很好的可扩充性,易于加入新的协作主机和入 侵检测Agent,也易于扩充新的入侵检测模式.它采用没有中心控制模块的并行Agent检测模式 ,各Agent之间的协作是通过它们之间的通信来完成的,各Ag |
| 关键词: 入侵检测,分布式模型,Agent,通信，协作. |
| DOI： |
| 分类号: |
| 基金项目:本文研究得到国家重点基础研究发展规划项目(No.G199035810)和中国科学院软件研 究所青年创新基金(No.cx2k5606)资助. |
|
| Distributed Model of Intrusion Detection System Based on Agent |
|
MA Heng-tai,JIANG Jian-chun,CHEN Wei-feng,QING Si-han
|
| Abstract: |
| The framework model proposed in this paper is a real time intrusion detection sy stem based on Agent, which provides an interface for intrusion detection com pone nts. Such interface can be used to detect intrusion behaviors based on both netw ork and hosts. According to the different system or network usage patterns and e nvironment diversity, a set of various agents will be created which cooperate to detect the anomalous aspects. The proposed model is an open system, which h as g ood scalability. It is easy to add new cooperating hosts and agents and to expan d new intrusion patterns. agents work in a concurrent way without any central co ntrolling module. The cooperation among Agents is implemented just by communicat ion. Agents are independent but are capable of communicating with each other whe n they take their actions. The state-checking and policy of authentication mech anism ensure the security of the agents themselves and the communication among t hem. This model is independent of specific application environment, thus providi ng a general-purpose framework for intrusion detection systems. |
| Key words: Intrusion detection, distributed model, Agent, communication, cooperation. |
