节点文献
基于系统调用的Linux系统入侵检测技术研究
Research on Syscall-based Intrusion Detection Technology for Linux System
【摘要】 提出了一种基于系统调用、面向进程的Linux系统入侵检测方法:利用LKM(Loadable Kernel Modules)技术在Linux内核空间获取检测源数据——所考察进程的系统调用,使用基于极大似然系统调用短序列的Markov模型提取进程的正常行为特征,据此识别进程的异常行为。通过实验表明了此方法的可行性和有效性;并分析了方法在实现中的关键问题。
【Abstract】 A process-orientated intrusion detection method based on system call for Linux system was proposed: The LKM(Loadable Kernel Modules) technology running in the kernel space was used to obtain the system calls of a process,and the Markov model based on the maximum likelihood syscall short sequences was used to describe the normal profile and to detect the anomaly of a process.Experiments show good feasibility and validity of this scheme.Some key problems of the implementation for the scheme were discussed.
【关键词】 入侵检测;
系统调用;
可加载内核模块技术;
极大似然;
马尔可夫模型;
【Key words】 IDS; system call; LKM(loadable kernel modules); maximum likelihood; Markov model;
【Key words】 IDS; system call; LKM(loadable kernel modules); maximum likelihood; Markov model;
【基金】 天津市科技发展计划资助项目(0524200);天津市自然科学基金资助项目(033800611)
- 【文献出处】 计算机应用研究 ,Application Research of Computers , 编辑部邮箱 ,2007年04期
- 【分类号】TP393.08
- 【被引频次】12
- 【下载频次】279