Adversarial detection as a zero-sum game

We propose a new game theoretic approach to estimate a binary random variable based on a vector of sensor measurements that may be corrupted by an adversary. The problem is formulated as a zero-sum partial information game in which a detector attempts to minimize the probability of error and an attacker attempts to maximize this probability. Explicit mixed policies are computed using the matrix form of the game and exploiting sensor symmetry to reduce complexity and find closed-form solutions.