Time as a parameter of statistical ensemble

The notion of time is derived as a parameter of statistical ensemble representing the underlying system. Varying population numbers of microstates in statistical ensemble result in different expectation values corresponding to different times. We show a single parameter which equates to the notion of time is logarithm of the total number of microstates in statistical ensemble. We discuss the implications of proposed model for some topics of modern physics: Poincar'e recurrence theorem vs. Second Law of Thermodynamics, matter vs. anti-matter asymmetry of the universe, expansion of the universe, Big Bang.