Computer Virus Spreading On Bipartite Scale-free Networks

Scale-free network can describe a wide range variety of systems in nature andsociety, say, the Internet, the World-Wide-Web, wealth distribution network, sciencecollaboration networks. The discovery of the scale-free network impels the people todeeply research the topology and dynamics of the network.With the development of the science and the popularization of computers, people’slife and work are more and more inseparable from the Internet. But at the same time,computer virus come after it, and can cause huge economic loses for individuals andbusiness at any time. Thus researching the spreading of the computer virus is providedwith great practical significance. In the past several years, the modeling and analysis ofcomputer virus spreading have received considerable attention. The analysis procedureand the result of the research can help people have a more comprehensive understandingof the spreading of the computer virus, and provide guidance for making effectiveimmunization strategy to the computer virus.The main contributions of my thesis are as follows:1, Following the evolving rules of the one part BA scale-free network, we proposea evolving models on bipartite scale-free network based on strict deduction. Applyingcontinuum theory and rate equation approach, the connectivity distribution of thenetworks are analyzed.2, The SIS epidemiological model on bipartite scale-free network is studied. Weuse bipartite scale-free (SF) network model to simulate the computer networks. On thenetwork, we model the SIS spreading model. By using the rate equation approach, weanalyze the relationship between epidemic threshold and infection rates ofclient-to-server and server-to-client. We find that the ratio of infected client density toinfected server density is determined by the topology of the network and the proportionbetween infection rate of clients and servers. We prove this conclusion through thedissemination of model simulation.3, Based on above model, we build the immunization model. For the randomstrategy and node degree strategy, we both carry simulation and analyze the result. Inthe same network, we find random strategy, the node degree strategy and neighbor node strategy all can against the spreading of computer viruses, and the best is the neighborimmunization strategy.