Model Checking Network Transmission Intervention Policies

Network exists widely in nature and human society. The spread of infections, such as computer viruses,diseases and rumors, poses great challenges to economic and societal system. It is an important issue in the network transmission control to protect individuals by finding effective intervention policies. This paper focuses on the network transmission model affected by time and adjacent nodes. Based on existing static protection work, a further study on dynamic protection cases is conducted, and four kinds of analyses of intervention policies are summarized.The intervention goals are expressed as logical properties and an effective approach of searching for intervention policies via model checking is proposed. Furthermore, experiments on random and small-world networks are conducted, demonstrating the advantages and disadvantages of model checkers NuSMV and SPIN on solving intervention problems. The results indicate that NuSMV needs less memory, but more time to complete searching, for example three times are unfinished in total six trails because of time exceeding in the dynamic recoverable problem; SPIN needs less time, but more memory, for example one time is unfinished in total six trails because of memory exceeding in the static unrecoverable problem.