A Game-Theoretic View on Objective Function Attack and ITS Defense

The objective function of a cognitive network is the basis for optimizing its operation parameters. However, one kind of attack, named OFA (Objective Function Attack), could exploit some cognitive characters to disrupt the optimizing process, which will result in the operation parameters deviating from the optimal value. Existed defense strategies for OFA mainly focus on verifying the fluctuation ranges of those parameters in a central way, while the long response time caused by central processing is ignored. Moreover, those methods are incapable of coping with attacks with a large scope. By analyzing the interactive process between OFA and its defense, a defense method based on differential game is proposed in this paper. This method constructs a game model for automatic defense, in which an immediate response-defense strategy and a threat factor are used. Simulation results show that the proposed method has a shorter response time and better defense effect compared with other existing methods.