【摘要】 针对容忍入侵系统中自适应能力较差的问题,提出了一种基于自适应大数表决机制的容忍入侵系统模型。该模型采用了适当的复制技术和自适应大数表决技术,使系统即使在遭受到恶意攻志的情况下仍能安全、可靠地运行,并能有效地识别出有故障的复制品,从而达到了容忍入侵的目的。通过使用基于前几轮表决的历史记录,使得表决器表决出的输出值更准确、更可靠并能及时地移除掉有故障的复制品。分析表明,方案具有安全性强,可靠性高,系统性能好以及易于实现等优点,特别适应用于对系统安全性、可靠性、可生存性,以及系统效率要求较高的分布式环境中。更多还原

【Abstract】 Current intrusion tolerant systems have some limits on adaptability. To solve this problem, a novel intrusion tolerant architecture with adaptive majority voting mechanism is proposed, which can be used to mask the effects of intrusion, thus ensuring systems survivability. Both replication and adaptive majority voting are used in the model to provide correct and reliable services for the system continually and to detect the faulty replicas despite the existence of fault or intrusion, thus achieving the effect of intrusion tolerace. The output of current voting is produced by using the history records of previous several vothing, so the results are more correct and reliable than the traditional majority voting. Aanlysis shows that this method has many advantages over those availabe ones’ such as high security and reliability, reasonable performance and easy realization.更多还原