Distributed Attacks as Security Games

Due to the development of easy-to-use software, distributed denial of service attacks have gone from theoretical weapons to being in the hands of criminals in less than twenty years. This paper analyzes the threat of DDoS attacks by developing a two-sided multiplayer model of security in which attackers attempt to deny service and defenders attempt to protect against attacks. Key findings include the existence of protection and non-attack Nash equilibria among different configurations of contribution functions and numbers of players, and validation of this model against infection data. I found that strategic attackers launch attacks only if protection will not occur. Therefore, the threat of protection can be enough to deter an attacker, but as the number of attackers grows, this equilibrium becomes increasingly unstable.