Trusted Dynamic Measurement Based on Interactive Markov Chains

Trusted computing ensures trustworthiness of a platform through extending the trust boundary from the root to the whole platform. Trusted measurement is invoked before the trust boundary is extended from one entity to including another. Static measurement, which takes place at startup, cannot ensure runtime trustworthiness, and therefore dynamic trusted measurement is indispensable to guarantee a computer platform to run dependably. According to dependability, availability and security of information and behavior, targets of trusted measurement are established. In present schemes of dynamic trusted measurement, the measurement of functionality is focused on, whereas dependability cannot be guaranteed without the measurement of performance. Based on interactive Markov chains (IMC), the measurement of performance feature besides function feature is introduced. In the expected behavior description, the function expectation is described through a model of transition system and the performance expectation is described through relating path probability indicating dependability to the time expectation in which a certain specific behavior function is achieved. By comparing the runtime evidence of a platform with a specific expectation, trusted verification on a combination of functionality and performance is achieved. The trusted dynamic measurement model based on IMC ensures dependability in the feature of performance besides function and guarantees trustworthiness of a platform across the board.