Uncertainty in Intrusion Detection Signaling Games

Abstract

In the area of Intrusion Detection (ID) games are being played between potential attackers and the Intrusion Detection Systems (IDSs) that protect the target systems of several attacks. More than a few game models presented in the past have showed how much beneficial Game Theory could be when incorporating with ID. In this research work an ID game model is constructed and examined as a signaling game. First, we construct the ID signaling game in an extensive form by defining the corresponding payoffs. Next, we represent it in a payoff matrix as a normal form game. We examine then the solution of the game by removing the dominated strategies. Finally, we compute all the equilibria of the ID signaling game in pure and behavioral strategies. The results give valuable explanations about how ID games are being played, what are players’ choices and under which circumstances, and the amount of uncertainty an ID game bears from its start point until the end.