This paper developed two fairly simple mathematical models of computer virus propagation. The simplicity was achieved by assuming many parameters constant. The applicability of the analytic expressions obtained from the model is therefore limited.
To come up with a more correct and reliable model, this paper recommends that the number of assumptions be reduced. Among those that should be eliminated are the assumptions made about the infection and disinfection rates. These rates should not be held constant because as more computers get infected, both the infection and disinfection rate increases with time.
Better models can also be achieved by including more parameters. This paper recommends the inclusion of such parameters as (a) the type of computer network topology (b) the level vigilance of the computer user in keeping the antivirus updated, (c) the effectiveness of the metrics used by computerized organization in combating viral attacks, etc.
The application of SI and SIR epidemiological models to the propagation of the computer virus provided the analytic expressions that describe the behavior of infection as well as the disinfection. It was observed that in the absence of an antivirus, the SI model is telling us that at some determinable time, all computers shall become infected with a virus.
The SIR model is telling us that in the presence of an antivirus, the increasing trend of infection will be reversed at some future time. Later from such time, there will be a steady state for the infection at which time the epidemic can be considered to have ended.
This paper concludes by asserting the mathematical modeling is useful not only in the study of computer viruses but in any fields where the behavior is associated with either growth or decay. The usefulness of the model is in describing what happens to the subject being studied at a particular time in the future. This description of the future will guide us what action to take and decision to make now for a brighter tomorrow.