Modelling influence of Botnet features on effectiveness of DDoS attacks

Abstract

The number of coordinated and targeted Distributed Denial of Service (DDoS) attacks in the world used both as informa- tion warfare and as economic racket is increasing. It demonstrates the importance of DDoS attack modelling in order to increase understanding of DDoS nature success factors and to refine the protection mechanisms. Existing models concentrate on DDoS attack and victim properties and do not pay attention to Botnet properties, while Botnets become the major DDoS organization and management technology. We propose the DDoS attack model, which allows estimation of influence of Botnet size and agent allocation strategies on attack success probability. Modelling results of different Botnet agent allocation strategies have shown the dependence of DDoS success on attack power dynamics. The proposed DDoS attack model can be used for analysing probability of victim resistance to DDoS attack by modelling situations with different victim, attack and Botnet properties and by applying these results for optimal countermeasure planning.