基于粒计算的入侵检测系统研究

Research of Intrusion Detection System Model Based on Granular Computing

入侵检测系统作为保护网络系统安全的关键技术和重要手段,已经成为当前网络安全研究方面的热点。粒计算作为人工智能领域中的一种新理论和新方法,将其应用在入侵检测中具有重要的现实意义。为了讨论该方向上尚未解决的问题和发展趋势,采用了用等价关系来描述粒,用粒来描述概念的方法,并用实验证明了该模型能非常有效地将正常和异常的系统调用序列区分开,得到用粒计算的方法论可以解决入侵检测问题的结论。

The intrusion detection system as a key technology and important means to protect network security has become a hotspot. Granular computing as a kind of new idea and method in the artificial intelligence has important practical significance in intrusion detection. To discuss the direction of unresolving issues and trends of development, the methods of using equivalence relation to describe the granular and using granular to describe the concept are proposed. Besides, the experiment verifies that this model can separate the normal system call from the abnormal one effectively. The conclusion is that the methodology of granular computing can be used to solve the problem of intrusion detection.