【摘要】 针对工业人机界面资源有限、实时性要求高等特点,分别从入侵过滤、入侵检测、入侵处理3个部分设计了工业人机界面的信息安全策略。其中入侵过滤采用"白名单"机制,入侵检测采用基于模型的异常检测机制,模型包括任务模型、CPU利用率模型以及网络流量模型。该策略在基于LPC3250-Linux的嵌入式平台上进行了实现和相关实验,实验结果表明,该策略能有效地检测出测试数据集中的入侵行为并保持较低误报率,同时满足系统实时性要求。更多还原

【Abstract】 Due to the limited resources and the hard real-time requirements of HMI,a security strategy is proposed.including intrusion filter,intrusion detecting and processing.The "white list" mechanism is used in the intrusion filter and anomaly detection mechanism based on models is applied to the intrusion detecting.The models include task model,CPU-utilization model and network-flow model.The security strategy is verified on a LPC3250-Linux platform.The results show that the intrusion behaviors in standard data sets are effectively detected and the rate of false positives is low.And they also indicate that the real-time requirement of the system is fulfilled.更多还原