中文摘要：

针对当前入侵容忍系统需要过多人工干预和缺乏自适应性的问题，提出了一个基于自律计算的容侵系统模型．该模型借鉴自律计算的思想并引入反馈机制对网络连接的可信度进行实时分析，生成初始信度，并通过信度阈值的自主学习与调整实施动态的信度优化，实现对可疑连接的自律容忍．

英文摘要：

According to the problem that the current intrusion tolerance system requires too much manual interven- tion and lack of self-adaptability, this paper proposes a model of intrusion tolerance system based on autonomic com- puting. The model drawing on autonomic computing ideas and feedback mechanism can analyze the reliability of the network connection in real time and generate initial reliability, and though the reliability threshold＇s independent study and adjustment, implement the dynamic reliability optimization, and ultimately tolerate the suspicious connec- tions, achieve the goal of protecting the target system security.