A stochastic epidemiological model for the propagation of active worms considering the dynamicity of network topology

Abstract

Topology-aware active worms, which use topological scanning for finding their victims, are one of the most serious threats in the Internet. Peer-to-peer (P2P) networks and applications are suitable environments for the spread of topology-aware active worms. Several models for the propagation behavior of these threats exist in the literature. Discrete-time models are usually more accurate than the continuous ones due to the nature of worm propagation, which is inherently a discrete-time process. On the other hand, as the propagation of worms is a stochastic process, the stochastic models enable us to study the stochastic characteristics of worm propagation process and are definitely useful. To the best of our knowledge, no stochastic model for the topology-aware active worm propagation has been developed yet. Also, none of the existing models consider the dynamic changes of network topology during the spread of worms. It is important that the network topology be taken into account as a key parameter in the model and at the same time, complex computations should be avoided. These are two important goals of this work, which were not considered in the existing models. In this paper, we introduce a new stochastic and discrete-time model for topology-aware active worm propagation (abbreviated as STAWP). The STAWP model considers the dynamicity of network topology and the join and leave of hosts in a simple manner. We have also extended the existing topology logic matrix (TLM) simulative model in order to meet the goals of the STAWP model. Comparing the results of our experiments using this extended model (i.e., extended TLM or ETLM) with the STAWP model, shows that their behaviors are nearly the same, which can be used to validate both models. Using the STAWP model, we have investigated the impact of several parameters in topology-aware active worm propagation process, the results of which are also presented in this paper.