Infection dynamics on the Internet

In previous works, the connectivity of nodes in social networks such as the Internet has been shown to follow a scale-free distribution in which there is a larger probability of nodes with lower connectivity and a smaller probability of nodes with higher connectivity. This network structure facilitates communication but also aids in the propagation of viruses. In this work, solutions have been obtained for a dynamical mean-field equation that characterizes virus infections and growth in scale-free networks. In contrast to previous findings, a threshold condition has been found for the persistence of computer infections. The effect of connectivity-dependent growth and recovery rates is also reported. It has been found that it is possible to reduce the deleterious effects of viruses by preferentially discouraging growth and enhancing recovery in high-connectivity nodes. Significantly, a security “figure-of-merit” has been derived that will allow network administrators to sample their environment in real time and measure the risk relative to E-mail-borne threats.