A discounted stochastic game approach to intrusion detection in wireless ad hoc networks

Abstract

A non-cooperative non-zero-sum discounted stochastic game framework is considered in order to methodically formulate the intrusion detection problem in wireless ad hoc networks. Underlying game parameters and player objective functions are defined to characterize the security requirements of a wide range of wireless ad hoc network applications. The validity and effectiveness of the developed framework are verified by generating numerical instances of the game model and analyzing the equilibrium results. The game outcomes show that the proposed framework provides useful insights on: (1) the expected behavior of rational intruders; and (2) optimal strategies of the intrusion detection system to allocate its security resources in different operational modes of the network.