针对当前入侵容忍系统无自适应能力、缺乏定量分析等问题,提出了一个基于自律计算的自适应分级入侵容忍模型(adaptive hierarchy intrusion tolerance system based on autonomic computing,AHIT^AC)。采用功能分级的分层模式,AHIT^AC的关键模块涵盖了访问连接的信度评估、可疑信息的主动诱骗、应用服务的分级学习与系统功能的分类恢复;同时通过自主实现可信度阈值、系统服务分级和诱骗知识库的学习和自适应过程,AHIT^AC实现了对入侵、可疑信息的有效容忍,提高了目标网络的自我修复能力和自我优化能力。实验结果表明,加载了AHIT^AC的目标网络服务性能稳定,容忍性能良好。
Aiming at the absence of self-adaptation ability and quantitative analysis on existent intrusion tolerance system, this paper proposed an adaptive hierarchy intrusion tolerance system based on autonomic computing ( AHIT^AC ). Adopting hierarchy modes, the critical modules of AHIT^AC included confidence evaluation of accessing, active trapping on suspicious information, hierarchy study of applications and classed recovery of system function. By implementing the study and adaptive function of confidence threshold, service classification and trap repository, AHIT^AC implemented the tolerance on intrusion and suspicious information, improving the ability of self recovery and self optimization on object network. The simulation results show that the object network with AHIT^AC is stable and tolerant.