Modeling Of Virus Propagation

Network technology has evolved by leaps and bounds nowadays,and the growing popularity of the Internet has greatly changed the way we work and live.However,since the emergence of the Internet,cyber security has become an unavoidable issue,and now it has risen to the national strategic height.The attacks of computer viruses(viruses,for short)are the most typical kind of cyber security incidents.Antivirus software,patches and firewall are the main technical means of defending against viruses,which can weed out all viruses they can recognize that stay in individual electronic devices(nodes,for short).Unfortunately,these techniques seem powerless to the outbreak of a new virus.In order to effectively contain virus diffusion,it is necessary to understand the propagation laws of viruses,which may provide a theoretical basis for decision-making,as well as to use technical measures.The propagation dynamics of computer viruses is aimed at establishing dynamical models capturing the propagation behaviors of viruses by taking into full account the characteristics of viruses and various factors,especially countermeasures,that have significant impacts on viral spread,analyzing the propagation laws of viruses and working out effective strategies to suppress virus prevalence.(1)This paper starting from the perspective of mechanism modeling and data modeling,to improve the existing SIRS model in which the network topology changes is not considered,we propose an improved model by taking into account network topology changes.The threshold and the correlation between the threshold of this model and the topology were deduced by the Lyapunov stability theory.The disease will disappear ultimately when the system meets the threshold condition.Except that,we also proved the existence of the equilibrium point of endemics and the limiting conditions for stability of the equilibrium when the system does not meet the threshold condition,The numerical simulation results indicated that the theoretical conclusions are valid,and the SIRS model with network topology changes can better simulate the spread of actual epidemics than the existing SIRS model.(2)In view of the shortage of the warehouse model in depicting the spread of virus,combined with the incidence data of scarlet fever from 2012 to 2017,this paper respectively establishes an improved model based on approximate analytical solutions of SIS model,a scarlet fever epidemic prediction model based on NAR neural network,and a scarlet fever epidemic prediction model based on local weighted linear regression.Uses the established three models to predict scarlet fever incidence from January to June 2018,respectively,by calculating errors and variances of each model,it is concluded that the local weighted linear regression model can more accurately describe the epidemic trend of scarlet fever than the approximate analytical solution model,the improved model and the neural network model.