CORBA分布式对象应用中的免疫型入侵检测研究

Immune-based Intrusion Detection for CORBA Distributed Object Applications

借鉴生物体免疫学原理,提出一个CORBA分布式对象应用中的入侵检测模型。模型采用CORBA应用客户的请求序列作为检测特征,将客户实时运行中的请求短序列与训练阶段建立的Self数据库比较,从而发现入侵行为。原型系统的实验表明,入侵检测模型能够对CORBA分布式对象应用中的异常客户行为进行有效检测。

The paper presents an intrusion detection model in CORBA distributed object applications based on immunological principles, with signatures based on sequences of clients requests to CORBA objects. During live operation, each new signature is compared with the self database contains all of the signatures collected during the training period, and concentrated bursts of anomalies signal an attack. The experimental results show the models effectiveness in detecting anomaly clients in CORBA distributed object applications.