Cascading failures dynamics model based on the road impedance and conditional Markov state

In this paper, a new dynamic model, based on Markov state transfer theory, is proposed to analyze the dynamic process of cascading failures caused by load fluctuations in communication networks. The model innovatively borrows the concept of road impedance from the field of transportation, which is used as the weights of network edges to reflect the interaction dynamics among network nodes more accurately, so as to optimize the dynamic allocation of load. Moreover, an extended Gillespie method is adopted to randomly select faulty nodes and determine the time between faults. Finally, the effects of network topology and parameter settings on the propagation and severity of cascading faults are explored through simulation experiments on Barabási-Albert networks, Community networks, and verified on real Autonomous System networks.