Health Analysis Of Cyber Physical System Based On Reliability

Cyber Physical System(CPS)consists of two interdependent networks,i.e.,a cyber network embedding into a physical network.Although interdependence makes the system more intelligent,it increases the vulnerability and complexity of the system in the presence of failures,and even leads to cascading failures,which may cause the entire system to completely fragment.In this p aper,we consider the health status of the cyber physical system in the presence of initial point failures.The main conclusions are as follows:1.We propose(m,k)-survivability and(m,k)-health to evaluate the probability and the expectation that at least k nodes still operate in steady state with the maximum entropy as m nodes are attacked.And both reflect the capability of system to resist the failures.Further we proved that the health of system decreases with the increase of m or k.2.Due to the complexity of calculating(m,k)-health parameter,we propose a Monte Carlo(MC)algorithm,and illustrate an unbiased and uniform estimation of the exact value.The simulation results show the proposed two parameters describe the health of the system well and the error between the estimated value and the exact value obtained by the approximation algorithm is less than 1%,which has a good approximation effect.