Skip to content 
The spread of infectious diseases in urban regions is governed not only by local transmission mechanisms but also by intercity connectivity patterns. This study models disease dynamics as a stochastic network process. The aims are twofold: the first is to determine the probability of a city network being in a given state, and the second is to determine the expected number of infected cities over time. Within the susceptible–infected–susceptible (SIS) framework, we formulate master equations for binary-state dynamics on arbitrary static graphs and simulate transmission in a small-scale city network. To parameterize and validate the model, empirical data are incorporated. Daily COVID-19 reports are aggregated into weekly city-level statuses based on standardized public health criteria. The infection period is estimated via maximum likelihood estimation (MLE), from which the recovery rate is derived, while the infection rate is inferred from the relationship between transmission probability and network connectivity. Results show that under high transmission probabilities, both the probability of network-wide infection and the expected number of infected cities initially increase but eventually converge to a disease-free equilibrium. Network topology significantly affects infection persistence, with highly connected networks exhibiting greater infection levels.