A multiresolution approach for optimal defense against random attacks

Whether it be one security expert covering more systems or reducing total man-hours, there has always been a push to do more with less. Intuitively, we realize different systems need different levels of security. To aid in this effort, we develop multiresolution attacker/defender games by combining two game theoretic approaches: resource assignment and optimal response. We use the resource assignment game to determine the level of detail necessary to build the game needed to respond optimally to attacks. To aid in the selection of a resource assignment game and an optimal response game, we present considerations and survey numerous works. Further resource savings are possible when the optimal response games share features. Even though effort sharing between systems ought to be addressed during the resource-allocation game, we present both a linear effort sharing model and a method for solving post hoc. An illustrative example demonstrates the potential savings from our technique.