动态环境中的概率信息流安全

Probabilistic Information Flow Security in Dynamic Context

在概率安全进程代数的基础上,研究了概率系统在动态环境中的信息流安全,以解决可动态配置的恶意进程对系统带来的危害.同时,提出了动态环境中的概率互拟复合不可演绎性质,即持久P-PBNDC.使用敌意环境下的弱概率互拟和动态敌意环境等概念讨论了P-PBNDC是概率系统在动态环境中合适的信息流安全性质,该性质能够揭露出概率互拟复合不可演绎性质(PBNDC)所不能暴露的潜在的安全隐患.最后,证明了P-PBNDC和SBSPNI性质的一致性,在限制算子和前缀算子下的P-PBNDC具有复合性质,而在并行算子下不具有复合性质.

Based on probabilistic security process algebra, information flow security of probabilistic processes in dynamic context is investigated, which can eliminate the danger brought by malicious processes that can be configured dynamically, and the persistent probabilistic bisimulation nondeducibility on composition property in dynamic context, namely P PBNDC is presented. That P_ PBNDC is a proper information flow security property of probabilistic processes in dynamic context is discussed by using the notions of weak bisimulation in hostile contexts and dynamic context, which can expose the potential secure problem that PBNDC can not discover. Finally it is proved that P_ PBNDC is consistent with SBSPNI, and P_ PBNDC has composite property on the restriction and prefix operator but not on the parallel operator.