基于行为的可信动态度量的状态空间约简研究

Research on State Space Reduction of Behavior-Based Trusted Dynamic Measurement

针对复杂并发计算机系统行为可信的动态度量研究中,细粒度动态度量所引发的状态空间爆炸问题一直是研究的难点.文中基于并发理论研究复杂并发计算机系统行为可信问题,在保障度量可靠性的前提下对系统状态空间进行约简,即通过标记变迁系统模型描述行为系统,通过事件结构模型研究行为关系,依据行为关系对变迁系统中各条路径进行重构,合并重构路径中相同的路径,实现变迁关系集约简,缩小状态空间.通过上述方法缓解了状态空间爆炸,并且,根据约简后的状态空间得到面向行为的可信动态度量的行为预期,增加细粒度动态度量方法在复杂系统中应用的可行性.

In research on dynamic measurement of behavior trustworthiness of complex concurrent computer systems,how to solve state space explosion caused by fine grained dynamic measure-ment is a difficult point up to the present.Based on concurrency theory,we researched behavior trustworthiness of complex concurrent computer system and reduced the state space on the premise of ensuring reliability of measurement.We described behavior system through the model of LabeledTransition System and researched behavior relationship through the model of event structure,then we reconstructed all paths in the transition system according to the behavior relationship and combined like paths in order to reduce the set of transition relations and narrow the state space.The situation of state space explosion can be relaxed through the methods mentioned above andmeanwhile,the behavior expectation of trusted dynamic measurement can be obtained according to the reduced state space,which enhances the feasibility of the method of fine grained trusted dynamic measurement applied in a complex system.